attraction.lab

Do Rebound Relationships Work? The Only Experiments Never Put Anybody in One

Not one of the eight readable pages gives a sample size, and the two experiments on the subject never put anybody in a new relationship.

Research-backed writing on attraction, dating and relationships. Drafted by AI against a named bibliography, published only when every claim links to its source.
35 min read
Editorial illustration for: Do Rebound Relationships Work? The Only Experiments Never Put Anybody in One

Before the article. If you are reading this in the middle of a breakup and tonight is a bad one, you do not need a research summary to get through it. Find A Helpline, in the sources at the bottom, lists free and confidential services by country, and in the United States you can call or text 988 at any hour. If the relationship that ended involved being hurt, controlled or threatened — or if a new one is starting that way — the National Domestic Violence Hotline is 1-800-799-7233, or text START to 88788. The last section says which patterns are a question for a professional rather than a question about timing.

With that said. The question in the title was put to DuckDuckGo on 14 November 2026 and cross-checked on Bing. Ten results came back; eight of them would serve themselves to be read. Not one of the eight gives the sample size of any study it describes. One links to a DOI, and that link is for a claim about rumination rather than anything about rebounds. Two link a reader straight to the paywall of the one paper with "rebound relationships" in its title without naming its authors, its journal or its size.

What makes this subject unusual is that the first page of results is not wrong about the direction. The best page on it says plainly that the folk rule — that dating too soon is doomed — does not survive contact with the research, and that is correct. The problem is that nothing on the page can be checked, and that the claims doing the most work carry no citation at all.

Followed back to the documents, the subject turns out to be four empirical papers, two of which cannot be read from here, resting between them on a few hundred people, most of them undergraduates at two universities. The one paper with experiments in it — the only random assignment in the field — did not put anybody in a new relationship. It had single students read a magazine article containing invented statistics about how easy it is to find someone, and then measured how attached they still felt to their ex.

What follows is sixteen sections: what the results page contains, counted; where its central claim comes from; what the experiments actually did; the confound the genre never mentions and the authors' answer to it; the moderator that gets dropped; the two studies nobody cites; where the popular "it's the motive, not the timing" line actually comes from; the one construct in this field that is well measured; the follow-up from the same lab that found less; and the thing in the title that has never been measured at all.

It does not tell you how long to wait, because nothing in the research supports a figure in either direction. Our editorial policy explains why every empirical claim here carries a linked source.

#1

What the first page of results actually contains, counted

The question in the title was put to DuckDuckGo on 14 November 2026 and cross-checked on Bing, which returned the same page in first position and reported about 36,500 results. Ten results came back. Two of them are Reddit threads that would not serve themselves to an automated request — both returned an eight-kilobyte stub — so what follows is counted over the eight that did.

The eight are Lifehack, Psychology Today twice (a December 2023 piece and a June 2014 one), Simply Psychology, BetterHelp, Verywell Mind, Bonobology and a site called psychology.tips.

  • Zero of the eight give the sample size of any study they describe. Not one page contains a figure in the shape of a sample size anywhere in its visible text.
  • One of the eight links to a DOI: Simply Psychology, once — and the link is not for any claim about rebounds. It points at a 2008 paper on rumination, attached to a sentence about when to seek help.
  • Two link to the publisher's page for the one study with "rebound relationships" in its title — Lifehack to the abstract, Bonobology to the paywalled PDF — and neither names its authors, its journal or its size.
  • Two name the authors of that study. Simply Psychology and psychology.tips.
  • One names the journal it appeared in. The 2014 Psychology Today piece.
  • Three contain the string "et al" with word boundaries anywhere on the page.

The counting is done this way for a reason that has already gone wrong on this site four times. Searching visible text for "doi" finds the letters inside "doing"; searching for "et al" finds them inside "get along". The counts above are of `doi.org` inside href attributes and of "et al" with word boundaries, which is the only version worth reporting — and the same caution applies to the literature itself, as a later section shows.

A result of zero out of eight is not unusual in this genre. It came out the same way for micro-cheating, where ten pages explained the term and none gave a sample size for anything.

So the useful thing to do here is not to add a ninth opinion. It is to find out what has actually been measured, by whom, on how many people, and with what design. That turns out to be a short list, and the most interesting thing on it is an experiment that did not test what its reputation says it tested.

#2

Credit where it is due: one page gets the direction right

This genre usually goes wrong by repeating a number that nobody checked. Here it goes wrong in the opposite direction, and the best page on the list is the reason this section exists.

Simply Psychology published a piece on 30 April 2026 that opens by saying the rebound rule "survives contact with the research" poorly, and that is correct. It states that quicker re-partnering has been associated with better wellbeing, more confidence and faster detachment from the ex, not worse. On the direction of the evidence, it is right, and it is close to alone in being right.

That is unusual enough to say plainly, because the premise this article was expected to have — that the whole genre insists rebounds are doomed and the research says otherwise — is not true of the first page of results. The strongest page already says the research says otherwise.

What it does not do is let a reader check anything. The page carries its own label: "AI-assisted draft · source-checked · editorial spot review", and a four-minute read time. For the central claim it names two surnames and stops. There is no journal, no year, no sample size, no link. And three of its load-bearing sentences carry no citation at all:

  • that "follow-up work on rebound timing has repeatedly failed to find the doom effect"
  • that research on ex-focused rebounds "finds exactly what you'd predict"
  • that rebounds entered for avoidance reasons "show poorer quality and stability" than those entered for approach reasons

The third of those is presented as the literature's real warning, the thing that matters more than timing. A later section follows it back to the nearest study that measured anything like it, and that study measured something else.

None of this makes the page's direction wrong. It makes the page unfalsifiable, which is a different problem and in some ways a worse one: a reader who wanted to know how much weight the claim can carry has nothing to weigh. It is also how claims survive after they stop being true — the most-repeated figures about what age gaps actually predict were withdrawn by the researchers who calculated them, and the pages repeating them had no mechanism for noticing.

The rest of this article is the weighing.

#3

The study everyone is pointing at, and what its abstract does and does not say

There is one paper with "rebound relationships" in its title that set out to test the belief directly. It is Claudia Chloe Brumbaugh and R. Chris Fraley, "Too fast, too soon? An empirical investigation into rebound relationships", in the Journal of Social and Personal Relationships, volume 32, issue 1, pages 99 to 118, published online on 5 March 2014.

Its abstract is worth quoting closely, because the gap between what it says and what it is said to say is where most of the trouble lives. It reports two studies of people who had experienced a breakup, assessing "their well-being, their feelings about their ex-partner, and whether they were seeing someone new". The findings, in the abstract's own words: people in new relationships "were more confident in their desirability and had more resolution over their ex-partner", and among those already in new relationships, the speed with which they began was associated with "greater psychological and relational health". The authors conclude that rebound relationships "may be more beneficial than typically believed".

Three things in that paragraph deserve attention. The comparison that produced the headline is between people who are in a new relationship and people who are not. The speed finding is reported within the group who already had a new partner, not between the fast and the patient. And the outcomes are confidence, resolution and self-reported health — not whether the new relationship went anywhere.

And then the position has to be stated plainly, because it decides what this page can responsibly do. The paper is paywalled. OpenAlex reports no open location for it; Semantic Scholar marks it CLOSED. The full texts that can be found are on a piracy mirror and a document-upload site, and those are not used here. So no number from inside this paper appears anywhere on this page — not a sample size, not an effect size, not a confidence interval.

Which produces the odd situation this article is really about. Two of the eight pages link a reader straight to that paywall. None tells them how big the study is. And neither, honestly, can this one.

This keeps happening, and it is worth knowing that the open-access labels are not reliable. The study behind the ick is indexed as open access by three separate services and returns a 403 on all three routes, and a paper in a later section here carries a CC BY licence and still refused to be read.

#4

Where its participants came from, and why that bounds the answer

One thing about the sample can be established without reading the paper, and it matters more than it sounds.

The paper's method section states that its participants came from a separate, larger longitudinal study of attachment in romantic relationships, citing Fraley, Vicary, Brumbaugh and Roisman (2011). That sentence is visible in the search engine's index of the paper's own text, and it is the one claim in this article that rests on a search index rather than on a document that could be opened. It is flagged as such deliberately.

The parent study, though, is fully readable, because Fraley hosts his papers on his own lab site. It is "Patterns of Stability in Adult Attachment: An Empirical Test of Two Models of Continuity and Change", in the Journal of Personality and Social Psychology, volume 101, issue 5, pages 974 to 992. It describes two non-overlapping samples, both recruited at a large midwestern university and the community around it:

  • a daily sample of 203 people, surveyed every day for 30 days, average age 21, 79% Caucasian;
  • a weekly sample of 388 people, surveyed weekly across a year, 65% female, average age 22.5.

Whatever subset of those people the rebound analyses used, that is the frame it came from: undergraduates and near-undergraduates at one American university, in their late teens and early twenties, already in a relationship when they were recruited, and then followed while some of them broke up.

That frame is genuinely good for this question — it is the reason the study can see a breakup happen rather than ask someone to remember one, which is the same design advantage that makes the only prospective study of recovery time worth reading in how long it actually takes to get over a breakup. It is also narrow in exactly the ways that matter to most people asking the question. Nobody here was married. Nobody was leaving a relationship of fifteen years. The average participant was twenty-two.

None of the eight pages mentions any of this. Two of them link to the paywall and call it research.

#5

The question does have randomized experiments, and nobody mentions them

Here is the part that changes the shape of the whole subject, and it took one search to find.

Everything above is correlational: people who have a new partner compared with people who do not. That design cannot tell you whether the new relationship did anything, because the people in it chose to be in it. The obvious next question is whether anyone has run an experiment — and the answer is yes, sixteen years ago, in a paper whose title is literally about rebounds.

It is Stephanie S. Spielmann, Geoff MacDonald and Anne E. Wilson, "On the Rebound: Focusing on Someone New Helps Anxiously Attached Individuals Let Go of Ex-Partners", in Personality and Social Psychology Bulletin, volume 35, issue 10, pages 1382 to 1394, 2009. Its abstract describes one correlational study and two experiments, in which detachment from an ex was "induced by randomly assigning anxiously attached individuals to believe they will easily find a new partner".

Zero of the eight pages mentions that experiments on this question exist.

That omission has a precedent on this site. The pages answering what to do when a partner refuses couples therapy gave advice and no numbers, while a randomized trial built for exactly that situation sat uncited. When a results page offers only guidance, the question to ask is not what they failed to cite. It is whether somebody ran a trial.

The published paper is paywalled. But there is an open route, and it is the same one that worked for the recent attachment-decay paper: the studies were Spielmann's Master's thesis at the University of Toronto, 2008, deposited in that university's institutional repository and readable in full. The thesis carries the same three studies under a slightly different title — "helps insecure individuals let go of ex-partners", where the journal version says "anxiously attached" — and every number in the next two sections comes from the thesis, not from the published article.

That distinction is worth keeping. The published version is the one the field cites, and it could not be read from here, so there is no way to confirm that the final sample sizes and coefficients match the thesis after peer review. Where this page gives a number for these studies, it is the thesis's number, and it says so.

With that said, what the experiments actually did is the single most useful thing on this page, and it is not what the reputation of the paper suggests.

#6

What the two experiments actually manipulated: a magazine article and a memory task

Neither experiment put anybody in a relationship. This is the point on which the entire popular reading of this literature turns.

Study 2. Ninety University of Toronto undergraduates, all single at the time of the session, took part for course credit; seven were excluded for having no experience of a breakup or for suspecting the task, leaving 83 in the analyses, average age 19.5. Each read one of two articles presented as coming from a relationship magazine. The optimistic version offered invented statistics — "a whopping 87% — reported being happier and more satisfied with their new partners", and that people "tended to find these new rewarding relationships within only five months of their breakups". The pessimistic version said the opposite, including that "nearly 87% of the respondents said that there simply 'wasn't anyone good to date'". Then they rated continued emotional attachment to their ex.

Study 3. Ninety-three undergraduates, again all single; ten excluded, leaving 83, average age 18.9. This time the manipulation was an ease-of-retrieval task: name two people you could plausibly date (easy, so finding someone feels easy) or ten (hard, so it feels difficult). Then the same attachment-to-ex measure.

So the experimental evidence in this literature is evidence about a belief. What was randomly assigned was optimism about how easily a new partner could be found. Nobody was assigned a new partner, a date, or a conversation with a stranger. The participants went home as single as they arrived.

That is a real and interesting finding — it suggests that some of what a new relationship is doing for someone may be done by the mere expectation that one is available, which is a claim about reassurance rather than about romance. It is also emphatically not a test of whether starting a relationship quickly after a breakup is good for you. The one place in this field where random assignment exists is the one place where rebound relationships are absent.

A genre that says "studies show rebounds help you move on" is describing an experiment in which the independent variable was a magazine article.

#7

The confound in the correlational study, and the authors' own answer to it

The first of Spielmann's three studies is the one that looks most like the claim people want, and it carries a problem that the thesis reports in its own results section.

The design. 162 University of Toronto introductory-psychology students who had been through a breakup took part. Thirteen were excluded, and the reasons are listed: four had no prior relationship experience, six had got back together with the ex, two had broken up more than ten years earlier, and one did not answer the questionnaire properly. That leaves 149 analysed — 69 currently in a new relationship (14 men, 55 women) and 80 single (22 men, 58 women). Ages ran from 17 to 34, averaging 19.7.

The problem. Those two groups were not the same distance from their breakups. The single participants had broken up between two weeks and 24 months earlier, averaging 7.5 months. The ones with a new partner had broken up between 1 and 66 months earlier, averaging 19.5 months — and the difference is large and significant, t(145) = −6.37, p < .001.

Read that again against what the study is used to prove. The group with new partners had, on average, been broken up more than twice as long. So this comparison is not between people who moved on fast and people who took their time. It is between people who currently have a partner and people who do not, and the ones who do are further from the ending.

The authors' answer, which deserves equal billing. The thesis addresses it directly: time since the breakup did not account for significant variance in attachment to an ex, either as a main effect or in interaction with the other variables, and including it did not change the pattern of results. Retrospective reports of how bad the breakup felt did not interact with anything either.

That is a reasonable response and it is more than most papers would offer. It is not the same as having matched the groups, and a reader deciding how much weight to put on the finding should know both halves. Not one of the eight pages reports either.

The general lesson is one this site has run into before: the question that moves a result is often not what was measured but which clock was started. In the literature on whether living together before marriage leads to divorce, the same dataset with the same controls produces a significant association or no association at all depending on whether age is counted from the wedding or from moving in.

#8

The moderator the genre drops: this was never a finding about everybody

There is a word in the title of the published paper that does not survive the trip into advice, and dropping it changes the claim completely.

The title is "Focusing on Someone New Helps Anxiously Attached Individuals Let Go of Ex-Partners". The effect is a moderation: the link between attachment anxiety and longing for an ex was disrupted when anxiously attached people had new partners, and the experiments induced detachment specifically in anxiously attached participants. The thesis frames the hypotheses the same way, around participants whose anxiety scores fell high rather than low.

For people low in attachment anxiety there was no comparable benefit to report, because there was less to disrupt: the longing the intervention relieved is the longing that attachment anxiety predicts in the first place. The finding is not "new relationships help people get over exes". It is closer to "for people whose insecurity keeps them attached to an ex, believing someone else is available loosens that grip".

Attachment anxiety is one of the two dimensions scored by our attachment style test, which is an educational tool and not a diagnosis or a validated clinical instrument. It is also not a fixed trait, which matters for how much a result like this can be generalised across a life — whether attachment styles change, and by how much has a measured answer, and that page is also where Brumbaugh's other work on this site appears.

The confusion this moderator prevents is a common one. Intense longing for an ex feels like a measure of how much the relationship mattered, and it is partly a measure of something else — how easily intensity and attachment anxiety get mistaken for love is the same problem seen from the other end.

So when a page says the research shows rebounds help you detach, the honest footnote is: in these studies, for people high in attachment anxiety, with a belief rather than a relationship as the manipulation, among Toronto undergraduates with an average age of nineteen.

Further illustration for: Do Rebound Relationships Work? The Only Experiments Never Put Anybody in One
#9

The second empirical study with rebounds in its title, which nobody cites

There is another paper that set out to study rebound relationships directly, and in eleven years the pages ranking for this question have not noticed it.

It is Cassie Shimek and Richard Bello, "Coping with Break-Ups: Rebound Relationships and Gender Socialization", in Social Sciences, volume 3, issue 1, pages 24 to 43, 2014. It studied 201 participants. Its finding, in the abstract's terms, is about who enters these relationships rather than what they do: men were predicted and found to be more likely to enter a rebound relationship after a breakup, and the predictors were lower levels of social support, more emotional attachment to the ex-partner, and a ludus, or game-playing, love style. The authors' summary is that rebound relationships functioned for these men as a distraction from continued attachment to an ex, and also as a substitute source of support.

Two notes on how it is cited here. It is published under a CC BY licence, which should make it fully readable — and the publisher's edge server refused every automated request from this machine, with a 403, including through a real browser. So this is the third time in a fortnight that a paper indexed as open access could not actually be opened, and like the others it is described from its published abstract only, with no figure taken from inside it.

The second note is the authors' own aside, which is the most quotable line in the literature: rebound relationships, they write, are "under studied by empirical researchers". That was 2014. It is still true.

The gender result is reported here because it is what the study measured, not as a claim about men in general: 201 participants in one study, with no replication, is a starting point and not a conclusion. But it points at something the advice pages do not consider at all. If entering a rebound tracks low social support, then the relationship may be doing the job that friends were not available to do, which makes it less a romantic decision than a resource one.

The ludus finding points somewhere else again, at a stable disposition rather than a reaction to the breakup. The nearest well-measured version of that idea on this site is the early signs of emophilia — the tendency to attach quickly and intensely regardless of who the person is, which is a trait rather than a stage of recovery.

#10

Rebound sex is a separate literature, and it is the one with real numbers

The strongest-designed study in this whole area is not about relationships. It is about sex, and it is the only one here that followed people forward through time.

Lindsay L. Barber and M. Lynne Cooper, "Rebound sex: Sexual motives and behaviors following a relationship breakup", in Archives of Sexual Behavior, volume 43, issue 2, pages 251 to 265, 2014. The method is a longitudinal online diary of 170 undergraduate students, tracking both psychological recovery and sexual experience after a breakup — so unlike every other study on this page, it watches the same people change over weeks rather than comparing different people at one moment.

What it found, from the abstract:

  • Having sex to cope with distress and to get over or get back at the ex-partner was elevated immediately after the breakup and then declined over time, as did the probability of having sex with a new partner.
  • People who had been "dumped" were more distressed and angry, and more likely to have sex to cope and to get back at or get over the ex.
  • People who reported coping or getting-over motives at the start were more likely to have sex with a stranger, and to keep having sex with new partners over time.

That last point is the only prospective evidence in this field that a motive at the beginning predicts behaviour later. Note carefully what the outcome is: more sexual partners over time. Not worse wellbeing, not a failed relationship, not a longer recovery. The study does not measure whether anything went badly.

The authors' own framing of the subject is the same as everyone else's who has actually worked on it — they discuss their results in terms of "widely held but largely untested beliefs about rebound and revenge sex".

It is worth separating this from the question in the title, because the two get merged constantly. Having sex after a breakup and starting a relationship after a breakup are different behaviours with different evidence, and only one of them has been followed over time. The territory between the two — sleeping with someone regularly without either of you defining it — has its own literature and its own measured problem, which is the subject of what the research says about situationships.

#11

Where "the motive matters more than the timing" actually comes from

The most repeated substantive claim on the best page in the results is that the real risk is not speed but motive: rebounds entered to escape pain, or out of anger, or to prove something, show poorer quality and stability than those entered out of genuine interest in the person. It is stated as the literature's considered warning. It carries no citation.

Followed back as far as it can be followed, the nearest measured thing is the study in the section above — and it does not say that. Barber and Cooper measured sexual motives and then sexual behaviour: how many partners, whether a stranger, how the probability changed over weeks. They did not measure the quality of a relationship, and they did not measure whether one lasted. A motive-to-behaviour finding about sex has been carried across into a motive-to-outcome claim about relationships, and the outcome in the claim was never an outcome in the study.

There is a real and large literature on approach and avoidance motivation in relationships, and there is a real finding in Shimek and Bello that a game-playing love style predicts entering a rebound. Neither of those is a test of whether avoidance-motivated new relationships turn out worse.

So the honest statement is a negative one, and it is stated with its limits: across the searches run for this article — the citation network of the two central papers, PubMed, Europe PMC and OpenAlex — no study was found that tests the quality or stability of a post-breakup relationship against the motive for entering it. If such a study exists, none of the eight pages cites it, and these searches did not surface it. That is a weaker claim than "it does not exist", and it is the strongest one the evidence supports.

This is the same failure the site keeps finding under different names: a plausible mechanism acquires the authority of a measurement by being stated next to real research. It is how the number attached to the 3-6-9 rule and the three-day texting rule got their reputations, and claims this site has had to withdraw are on the corrections page.

#12

Fear of being single: the construct with the clearest numbers in the field

If any part of this subject is well measured, it is not the rebound. It is the thing that makes an ex hard to let go of, and the construct has a name and two samples behind it.

Stephanie S. Spielmann, Geoff MacDonald, Samantha Joel and Emily A. Impett, "Longing for Ex-Partners out of Fear of Being Single", in the Journal of Personality, volume 84, issue 6, pages 799 to 808, published online on 29 September 2015. Its sample sizes are in the published abstract, so they can be given here without reading the paper:

  • Study 1: a cross-sectional study of 209 people, 64% women, average age 30 — notably older than the undergraduate samples above.
  • Study 2: a one-month daily-experience study of 117 people, 44% women, average age 27, all of whom had just been through a breakup.

The findings run in a consistent direction. People with a stronger fear of being single reported greater longing for their ex-partners in both studies. Fear of being single rose after the breakup, regardless of who had ended it. Within days, longing for the ex and attempts to renew the relationship were higher on days when that fear was stronger.

And the part that does real work, because it is the only place in this literature with a usable claim about direction: lagged-day analyses supported fear of being single increasing longing and renewal attempts over time, while longing and renewal attempts did not increase fear of being single. The arrow, in this dataset, runs from the fear to the longing.

That reframes the whole conversation. The question people ask is about the new person and the timing. The measured driver here is about being single, and it predicts reaching backwards — texting the ex, trying to restart it — rather than forwards. It is the same motive the research behind the no-contact rule keeps running into, and it is closely related to the pull described in staying friends with an ex and to why people stay in relationships that are not making them happy.

#13

The follow-up from the same lab that nobody has connected

The 2015 rebound paper's most quoted result is that people in new relationships had more resolution over their ex-partner. Eleven years later, the same senior author published a much larger study that looked at something very close to that, and did not find it.

The paper is Jia Y. Chong and R. Chris Fraley, on the long-term stability of affective bonds after romantic separation, in Social Psychological and Personality Science, 2025 — the paper behind the widely repeated "it takes four years to get over an ex" figure, which is taken apart in detail in how long it takes to get over a breakup. Its published version is paywalled; the openly available 2022 dissertation behind it carries the same sample and tables.

320 people, average age 32, each asked how much they would still want to turn to a former partner for the classic attachment functions. Among them, 58% had been in a committed relationship with somebody new since the breakup — and that was associated with essentially nothing. No main effect, no interaction with time, one small exception on one composite. Whatever a new relationship did for these people, it did not shorten the bond to the previous one.

Two cautions, because the temptation is to call this a failed replication and it is not quite one. The constructs differ: "resolution over an ex-partner" and "would still want this person as a secure base" are related but not the same measure. And both studies are cross-sectional, so neither can establish what caused what. This is the nearest thing to a check on the field's central claim, not a refutation of it.

But the tension is real, it involves the same laboratory, the larger sample is the one that found less, and zero of the eight pages cites the 2025 paper at all — including the page that asserts, without a source, that follow-up work has repeatedly failed to find the doom effect. The follow-up that exists found something more awkward than that: it failed to find the benefit.

#14

The thing in the title has never been measured

Step back from all of it and notice what is missing. The question is whether rebound relationships work. Every study above measures the person — their wellbeing, their confidence, their attachment to the ex, their sexual behaviour. Not one of them measures the relationship.

Nobody has followed rebound relationships forward to see whether they survive. There is no study here of how long they last, no comparison of break-up rates between relationships started quickly and slowly, no measure of satisfaction inside them a year on. A PubMed search for rebound relationship duration and for re-partnering speed against subsequent relationship quality returns nothing on the subject. The Brumbaugh and Fraley abstract reports psychological and relational health among people in new relationships, which is a self-report at one moment, not an outcome of the relationship over time.

This is why the pages promising to tell you how long rebound relationships last can only tell you stages and signs. There is no measured figure, because the measurement was never taken.

The nearest authoritative statement comes from the one systematic review in the neighbourhood: Joana Fernandes, Carina Santos and Mariana Veloso Martins reviewed the psychosocial effects of romantic breakups in emerging adulthood in Clinical Psychology & Psychotherapy in 2025, following PRISMA across three databases, with 47 studies eligible. Its conclusion on the point that matters here is that the long-term effects remain inconclusive and that further research is needed.

So the honest answer to the title is in three parts. Do people who re-partner quickly report doing worse? No — the evidence, thin as it is, points the other way. Does starting a relationship quickly cause anyone to do better? Unknown, because the only random assignment in this field assigned a magazine article. Do these relationships last? Nobody has looked.

That is a less satisfying answer than a number, and it is the one the research supports. The same gap shape — a confident genre sitting on top of an unmeasured question — turns up in how many dates it takes to know.

#15

The whole literature, counted, and the follow-up that never came

It is worth seeing how small this field is, because the confidence of the advice is inversely proportional to it.

Brumbaugh and Fraley's paper has been cited 47 times in eleven years, by the count of the citation database used here. Going through that list, something stands out: neither Brumbaugh nor Fraley ever returned to rebound relationships. There is no follow-up study, no replication, no extension by the people who did the original work. For a finding this counterintuitive and this widely repeated, that is unusual, and it is the opposite of what happened with the last two subjects this site looked at, where the authors' own follow-ups were the thing the search results had missed.

What the citing literature does contain is mostly borrowings: breakup reviews, a chapter on affective self-regulation, dissertations on repartnering, a study on mothers' repartnering after divorce. The nearest non-North-American contribution is a 2019 paper in the Japanese Journal of Psychology on attachment needs toward an ex delaying the transition through dissolution stages — openly readable, and cited once.

And a counting note, because it is the fourth time in a fortnight this has mattered. Searching a bibliographic database for the phrase "rebound relationship" returns 108 records, which sounds like a literature. Read them and they are mostly not: HIV-1 viral rebound dynamics, a 1962 paper on a strength criterion for repeated metal loads, the China–Gulf Co-operation Council relationship, corporate alliances after divestitures, and two separate analyses of figurative language in Ariana Grande lyrics. A substring is not a subject.

Stripped down, the empirical core of this topic is four papers: Brumbaugh and Fraley on wellbeing, Spielmann and colleagues on letting go, Shimek and Bello on who enters one, Barber and Cooper on sex. Two of the four cannot be read from here. Between them they rest on a few hundred people, most of them undergraduates at two universities.

A thin literature is not the same as a wrong one, and it is better than the alternative failure. When every page in a subject repeats the same confident figure, the thing to check is where the figure came from — in the case of whether opposites attract, the number the entire results page quotes turned out to be a sentence from a university press release rather than anything in the paper.

#16

What none of this can tell you, and when it is not a timing question

Put the samples side by side and the limits are obvious. Three studies of Toronto undergraduates, 315 people in total, average ages between 18.9 and 19.7, two of them manipulating a belief rather than a relationship. 201 participants in a single unreplicated study of who enters a rebound. 170 undergraduates in the one diary study, which was about sex. 320 internet and undergraduate participants, surveyed once, in the 2025 bond study. And the paper the whole subject rests on, whose size this page cannot tell you because it could not be read.

Almost none of this is about marriages. Almost none is about same-sex couples. None is about a relationship of twenty years ending at fifty, which is the situation many people asking this question are in. Nothing in it supports a waiting period of any length — and nothing in it supports hurrying either, because the one place in the field with random assignment never put anybody in a relationship.

What survives is modest. People who re-partner sooner do not, in the available data, report doing worse. For people high in attachment anxiety, a sense that someone else is available appears to loosen the attachment to an ex. Fear of being single predicts reaching backwards rather than forwards. And whether the new relationship itself lasts is not known to anyone.

Two patterns are different in kind from timing, and they are the reason the last section is this one.

The first is about distress that is not moving: not fluctuation, which is expected and normal, but flatness across months, or distress getting steadily worse, or any thought of harming yourself. That is not a question about how soon you started dating again, and the markers of it are covered in what has actually been measured about having moved on and, read from before an ending rather than after, in the longitudinal data on thinking about breaking up.

The second is about the new relationship, and it needs naming plainly. If someone new monitors who you see or where you are, pressures you to cut contact with friends or family, needs to know what you are doing at all times, or makes you feel responsible for managing their anger — those are warning signs of control, and they are not a rebound question. The same is true in the other direction: a relationship being conducted at an ex, to punish or provoke them, is unfair to the person in it, and that is a judgement about treating people decently rather than a finding from a study.

In the United States you can call or text 988 at any hour, and the National Domestic Violence Hotline is 1-800-799-7233 or text START to 88788. Find A Helpline, in the sources below, lists free and confidential services by country. The rest of this section is under Mindset.

Pulling it together

If one thing survives, it is the distinction between a direction and a measurement. The direction of the evidence on rebound relationships is genuinely reassuring, and it is the opposite of the folk rule: people who start something new sooner do not report doing worse, and for those whose insecurity keeps them tied to an ex, a sense that somebody else is out there appears to loosen that tie. Nothing here supports a mandatory period of solitude.

But almost none of it is a measurement of what the question asks. The only random assignment in the field manipulated a belief with a fabricated magazine article, not a relationship. The comparison behind the headline is between people who have a partner and people who do not, and in the study most often invoked, the ones with partners had been broken up more than twice as long. Whether these relationships last has never been studied by anyone.

So the useful conclusion is not a timeline. It is that nobody has earned the authority to tell you that you are moving too fast — and that the questions worth asking about a new relationship are about how you are being treated in it, which is not a matter of the calendar at all.

Frequently asked questions

Do rebound relationships work?

There is no measured answer to the question as asked, because nobody has followed rebound relationships forward to see whether they last. What has been measured is how the person does, and that evidence points against the folk rule rather than for it. Brumbaugh and Fraley, in the Journal of Social and Personal Relationships, reported across two studies that people in new relationships after a breakup were more confident in their desirability and had more resolution about their ex-partner, and that among those already in a new relationship, having started it sooner went with better self-reported psychological and relational health. That paper is paywalled, so no figure from inside it appears here, and none of the eight readable pages ranking for this question gives its sample size. What is missing from the whole field is any outcome measured on the relationship itself: no survival rates, no satisfaction a year later, no comparison of break-up rates between fast and slow starts.

Is there any experiment showing that rebound relationships help?

There are two experiments, and this is the thing most worth knowing about them: neither put anybody in a relationship. In Spielmann, MacDonald and Wilson's 2009 paper in Personality and Social Psychology Bulletin, single undergraduates were randomly assigned to conditions designed to make finding a new partner feel easy or difficult — in one study by reading a magazine article with invented statistics, in the other by being asked to name either two or ten people they could plausibly date — and then rated how attached they still felt to their ex. The numbers given on this page come from the openly readable 2008 Master's thesis version at the University of Toronto, since the published article is paywalled: 83 participants analysed in each experiment. So the random assignment in this field was of optimism about availability, not of a rebound relationship. It is evidence that reassurance can loosen attachment to an ex, which is a narrower and more interesting claim.

How long should I wait before dating again after a breakup?

Nothing in this research supports a figure, in either direction, and the rules in circulation do not come from studies. The frequently repeated idea of one month per year of the relationship has no research behind it, in the same way that the half-the-relationship rule for recovery time traces to two television comedies rather than to any paper. Nor does the evidence support hurrying: the only experiments in the field never assigned anybody a relationship, so there is no causal finding that starting sooner produces anything. What the data do suggest is that people who re-partner sooner are not, on average, reporting worse outcomes, which removes the main reason people give for waiting. Beyond that, the measured risks are about who is in the relationship and how it is going rather than about when it started.

Does a new relationship help you get over your ex?

The two studies that looked disagree, and the larger one found less. Spielmann's correlational study of 149 Toronto undergraduates found that having a new partner disrupted the usual link between attachment anxiety and longing for an ex — but the participants with new partners had broken up an average of 19.5 months earlier against 7.5 months for the single group, a large and statistically significant difference, though the authors report that time since the breakup did not account for significant variance in the outcome. Eleven years later, Chong and Fraley asked 320 people whether they had been in a committed relationship with someone new since their breakup; 58% had, and it was associated with essentially nothing on how much they still wanted to turn to the ex for attachment functions. The constructs are not identical, so this is not a strict failed replication, but it is the nearest check the field has, and none of the pages ranking for this question mentions it.

Are rebound relationships unfair to the new person?

That is a question about conduct rather than about evidence, and it is worth keeping the two apart. No study has measured how partners in rebound relationships are treated or how they fare, so there is no finding to quote. What the research does say is that people enter these relationships for mixed reasons: Shimek and Bello, studying 201 participants, found entering one predicted by low social support, continued emotional attachment to the ex, and a game-playing love style, and Barber and Cooper found that coping and getting-over motives after a breakup were elevated at first and then declined. None of that tells you about a particular relationship. The plain point stands on its own without a citation: conducting a relationship at an ex — to be seen, to provoke, to settle something — is unfair to the person inside it, and that is a judgement about treating people decently rather than a research result.

Is it true that the motive matters more than the timing?

It is a plausible idea that is being presented as a finding, and the finding behind it is about something else. The claim in circulation is that relationships entered for avoidance reasons — escaping pain, anger at an ex, proving something — show poorer quality and stability than those entered out of genuine interest. Followed back, the nearest study is Barber and Cooper's diary study of 170 undergraduates, and it measured sexual motives and then sexual behaviour: people with coping or getting-over motives at the start were more likely to have sex with a stranger and to keep having sex with new partners over time. It did not measure relationship quality, and it did not measure whether any relationship lasted. Across the searches run for this article, no study was found that tests the quality or stability of a post-breakup relationship against the motive for entering it. That is weaker than saying none exists, and it is as far as the evidence goes.

How long do rebound relationships last?

Nobody has measured it. This is the clearest gap in the subject: every study available measures the person who had the breakup — their wellbeing, their confidence, their attachment to the ex, their sexual behaviour — and not one measures the new relationship's survival, satisfaction or duration. Searches of PubMed for rebound relationship duration and for re-partnering speed against subsequent relationship quality return nothing on the topic. The one systematic review nearby, Fernandes, Santos and Veloso Martins in Clinical Psychology and Psychotherapy in 2025, screened 47 eligible studies on the psychosocial effects of breakups in emerging adulthood and concluded that long-term effects remain inconclusive. So pages that promise to tell you how long a rebound lasts are describing stages and signs, not a measured duration, because the measurement has never been taken.

Am I someone's rebound, and does it matter?

There is no test for it in the research, and the honest version of the answer is about the present rather than about their history. No study has measured how to identify being someone's rebound, or what happens to people who are. What the evidence does establish is that continued attachment to an ex is common, declines slowly and is not in itself a sign that something has gone wrong — in one study of 320 people the wish to still be able to talk to a former partner sat above the level people report for a stranger even a year out. So a new partner having feelings about their ex is ordinary, and the useful questions are observable: are they interested in your life specifically, is the ex a constant presence in how the relationship is conducted, and are you being treated well. Those can be answered from how things actually go, which is more reliable than anything a study could tell you about the category.

When is a breakup something to see a professional about?

The signal is not how fast or slowly you are moving on. Recovery in this literature is volatile, particularly early, so a bad week after a good one is an expected feature rather than evidence of being stuck. The patterns worth attention are flatness — no movement at all across months — distress that is steadily worsening rather than fluctuating, an inability to function at work or with other people that is not easing, or any thought of harming yourself. Those are questions for a professional rather than questions about timing. Separately, if a relationship you are in involves someone monitoring where you are, pressuring you to cut contact with friends or family, or making you responsible for managing their anger, those are warning signs of control and not a matter of recovery at all. In the United States you can call or text 988 at any hour, and the National Domestic Violence Hotline is 1-800-799-7233 or text START to 88788.

Sources

What this article draws on. For the broader bibliography see our sources page.

Keep reading