How long does it take to stop wanting them back?

No study supports a timetable, and anyone quoting a specific number of months is not citing research. What the forecasting work suggests is narrower: the decline is roughly as fast as people expect, and it is the starting height they get wrong. Those participants were students in relationships of a few months, so it says nothing about the pace after a long marriage.

Is it a bad sign that I mostly feel relieved?

Relief is common and is not evidence of having been callous. The one study following life satisfaction across divorce found gains concentrated among women who had been in very low quality marriages, which is consistent with relief carrying real information about what was left. It is still worth expecting the feeling to move around rather than settle.

We got back together and it feels different this time. Does that mean it will last?

The research speaks to averages and not to your case, and the averages are not encouraging: renewed relationships report more uncertainty and lower satisfaction than ones that never split, and each additional cycle tracks with worse ratings. The pattern does suggest one question worth asking directly, which is whether the specific thing that ended it has changed, or whether the reunion is running on relief.

BreakupsModerate evidence

Do people regret breaking up?

The short answer

There is no reliable figure for how many people regret a breakup, and any percentage you have seen was not produced by research. What the evidence establishes is narrower: the pull to go back is common and strongest in the early weeks, when predictions about your own feelings are least accurate; people who split and reconcile report worse relationships afterwards on average, not better; and whether someone ends up glad depends heavily on what they left.

16 min readPublished July 28, 2026

What readers think

Thinking about the breakup that brought you here, where do you land today?

Anonymous. One vote per reader.

The question usually arrives at a specific hour, some weeks in, when the certainty that carried the decision has worn off and the reasons no longer feel as solid as they did when you were living them. It is worth saying at the outset that no study can tell you whether you personally made a mistake, and this page will not pretend otherwise. What research can do is describe the shape of what you are feeling, which turns out to be regular enough to have been measured, and separate the parts of it that predict something from the parts that predict nothing at all.

There is no regret rate, and any figure you have seen was invented

Nobody has run the study that would settle this, which would mean following a large representative sample through breakups and asking them years later whether they wished they had stayed. What exists measures things adjacent to regret. The closest useful data comes from a survey of 2,484 parents going through divorce, taken after they had filed and after they had sat through a mandatory parenting class[7]. Around one in four individuals still believed the marriage could be saved, and in roughly one in nine couples where both partners answered, both of them did. Interest in reconciliation was patterned rather than random: it ran considerably higher among people whose spouse had initiated the divorce than among those who had initiated it themselves. That is doubt at the threshold rather than regret afterwards, and the distinction matters, but it establishes something worth holding onto. Substantial numbers of people reach the formal end of a relationship still unconvinced, and they are disproportionately the ones who were left rather than the ones who left.

The pull to go back peaks exactly when your judgement is worst

The most directly relevant experiment tracked people in dating relationships every two weeks for thirty eight weeks, twenty sessions in all, and analysed the twenty six who broke up during the first six months[1]. Because the schedule was fixed in advance rather than triggered by the breakup, the researchers held a forecast of expected distress made roughly two weeks before it happened, which they could then set against what those same people reported afterwards. The predictions were far bleaker than the reality. The interesting part is where the error sat. People were not wrong about how fast the feeling would fade, since the rate of decline broadly matched what they had expected. They were wrong about the height it would start from. This is a specific instance of a general finding that people overestimate the intensity and duration of future emotional states, partly because they picture the loss in isolation and leave out everything else that will still be going on around it[8]. The practical consequence is uncomfortable rather than reassuring. The days when going back feels most self evidently correct are the days your own reading of the situation is furthest from what you will report a month later.

Going back is common, and on average it does not go better

If regret is measured by what people do rather than what they say, a great many people act on it. Across two studies of American college students, almost two thirds reported having been in a relationship that ended and then restarted at least once[2]. This is not confined to young dating. In nationally representative samples of 323 cohabiting and 752 married couples, more than a third of cohabiters and a fifth of spouses said they had broken up and got back together within their current relationship[3]. What the same research finds about those relationships is the part that rarely gets repeated. On-again partners reported fewer of the positives, including love and feeling understood, and more of the negatives, including communication problems and uncertainty about where they stood, than people who had never split. A greater number of renewals tracked with worse ratings, not better. The cohabiting and married data point the same way and add that people who had cycled reported more uncertainty about the relationship's future and more of the practical constraints that make leaving hard. None of this is causal evidence. Cycling may be a symptom of instability that was there anyway rather than something that damages a relationship by itself. But if the hoped-for cure for the regret is reunion, the average case does not deliver it.

Whether people end up glad depends on what they left

A study using two waves of a national American survey roughly ten years apart, covering 1,639 participants, looked at life satisfaction before and after divorce and found the outcome was not uniform[4]. Women who divorced out of very low quality marriages reported higher life satisfaction afterwards. Women who left marriages they had rated more positively did not show that gain. The moderation is the entire finding: the direction of the effect tracked the quality of the thing that ended. Two consequences follow. The first is that a single number for how many people regret a breakup would be close to meaningless even if someone had collected it, because the population it averages over is not one population. The second is more sobering than the usual framing allows. If leaving a poor relationship predicts feeling better, the same arithmetic implies people who left something that was not poor, and for them this research promises nothing. It is also worth stating precisely what was found: this pattern held for women, it did not hold in the same way for men in the same analysis, and the study cannot say why.

The lessons-learned story is real as an experience and shaky as a fact

A frequently cited study asked 92 undergraduates who had recently been through a breakup what had changed for them, and they named on average around five kinds of personal growth, mostly things they believed would improve their next relationship[5]. That finding has been repeated often enough to have hardened into advice. A later study went at it prospectively, which is the harder design: 599 students in relationships were measured, then measured again about ten weeks on, with the hundred who broke up compared against those who had not[6]. Reported growth was high. Measured change was not, and the authors concluded that the pattern fitted positive reappraisal, meaning a revised account of what happened, better than it fitted demonstrated development. This is worth knowing in both directions. It does not mean people are lying, and reappraisal is not a worthless thing to have, since it is most of what anyone means by making peace with something. It does mean that a confident narrative about how much you have grown is not independent evidence that you were right to leave, in exactly the way that a confident narrative about how much you lost is not evidence that you were wrong.

It depends, on what?

What changes the answer

Two limits sit on top of all of the above, and they are large. The first is who was studied. Almost all of this comes from young, mostly Western, mostly student samples in dating relationships lasting months. The forecasting study that anchors the reassuring part of the answer analysed twenty six people. Extending that to a fifteen-year marriage with children, a shared mortgage and two families who hold opinions about it is not a small extrapolation, and where researchers have looked at marriage the picture becomes more conditional rather than more comforting. The second limit is what averages do. A finding that most people cope better than they predicted is a statement about a distribution, and distributions contain the people for whom it was not true. Some people end relationships they should not have ended, and no study identifies them in advance or afterwards. If you are one of them, the average is not a rebuttal and nobody should hand it to you as one. It is also worth separating two things that feel identical from the inside. Grief is the response to a loss, including a loss you chose deliberately and would choose again. Regret is a judgement that the choice itself was wrong. They arrive together, and the first is routinely mistaken for the second, particularly in the early weeks when, as above, the reading is least reliable.

Where people disagree

Both sides of it

Reasonable people land in different places on this. Here is the strongest version of each case, not a straw man of the one we disagree with.

The case for yes

Large numbers of people act on the impulse to reverse a breakup. Nearly two thirds of one student sample had been through at least one split and reunion, and more than a third of cohabiting partners had done it inside their current relationship.
At the formal end of a marriage, around a quarter of divorcing parents still thought it could be saved, which means substantial doubt survives well past the point where the decision is supposed to be settled.
Regret has an asymmetry the reassuring research tends to skip. The person who was left made no decision at all, so what they carry is closer to loss than to second thoughts, and the reconciliation data show their interest in reversing it is markedly higher.
The evidence that people feel better than expected comes from short dating relationships. It has not been shown to hold where the thing that ended was long, legally entangled, or shared with children.

The case for no

Feelings recorded in the first weeks are a poor guide to anything, since forecasts made before a breakup run systematically too dark and the error concentrates in precisely that early period.
Wanting someone back is not the same as judging the decision wrong, and the two get reported as one thing largely because they feel like one thing.
Where reunions happen, the resulting relationships rate worse on average, which suggests the regret was often about the loss rather than about that specific person or those specific reasons.
Retrospective accounts of a breakup are heavily reconstructed. A prospective study with a comparison group found that reported growth looked more like revised interpretation than measured change, and there is no reason reported regret would be less revisable.

Evidence

What the research says

Moderate evidenceTwenty six participants who broke up, assessed every two weeks across thirty eight weeks
In a biweekly longitudinal study, people forecast substantially more distress following a breakup than they went on to report. The error was accounted for by overestimation of the initial intensity of distress rather than by any misjudgement of how quickly it would fade.
What this does not show
The analytic sample was twenty six undergraduates in dating relationships of at least two months, followed for under a year. Nothing in the design speaks to marriage, cohabitation or relationships involving children, and a sample this small supports the direction of the effect far better than its magnitude.
Eastwick, P. W., Finkel, E. J., Krishnamurti, T., & Loewenstein, G., Mispredicting distress following romantic breakup: Revealing the time course of the affective forecasting error (2008)(opens in a new tab)
Moderate evidence445 students in the first study and 236 in the second
Almost two thirds of surveyed college students had experienced a relationship that ended and restarted. Those in on-again relationships reported fewer positives such as love and feeling understood, and more negatives such as communication problems and uncertainty, than those who had never broken up, with a greater number of renewals associated with worse ratings.
What this does not show
Cross-sectional self-report from United States college students across two samples, so it describes association rather than showing that cycling causes the poorer quality. Relationships that cycle may have been less stable to begin with, which the design cannot rule out.
Dailey, R. M., Pfiester, A., Jin, B., Beck, G., & Clark, G., On-again/off-again dating relationships: How are they different from other dating relationships? (2009)(opens in a new tab)
Moderate evidence323 cohabiting couples and 752 married couples
In nationally representative samples of cohabiting and married couples, over a third of cohabiters and a fifth of spouses reported having broken up and renewed their current relationship. Cycling was associated with lower satisfaction, greater uncertainty about the relationship's future, and stronger constraints against ending it.
What this does not show
Correlational, and reliant on retrospective reports of breakups within the current relationship. It establishes that cycling carries into cohabitation and marriage and travels with poorer outcomes, not that it produces them.
Vennum, A., Lindstrom, R., Monk, J. K., & Adams, R., "It's complicated": The continuity and correlates of cycling in cohabiting and marital relationships (2014)(opens in a new tab)
Moderate evidence1,639 participants across two waves roughly ten years apart
Using two waves of a national survey about ten years apart, women who divorced out of very low quality marriages reported increased life satisfaction, while women leaving higher quality marriages did not show that gain.
What this does not show
Observational, with only two measurement points a decade apart, so the trajectory in between is unobserved and divorce was not randomly assigned. Marital quality was self-rated, the moderation was found for women rather than men in the same analysis, and the sample is American and middle-aged.
Bourassa, K. J., Sbarra, D. A., & Whisman, M. A., Women in very low quality marriages gain life satisfaction following divorce (2015)(opens in a new tab)
Limited evidence92 undergraduates
Undergraduates who had recently been through a breakup reported on average around five distinct positive changes they attributed to it, alongside elevated distress.
What this does not show
Retrospective self-report from 92 students with no pre-breakup measurement and no comparison group, so the reported changes cannot be separated from a revised account of events. Women reported more growth than men, and the sample is a single undergraduate population.
Tashiro, T., & Frazier, P., "I'll never be in a relationship like that again": Personal growth following romantic relationship breakups (2003)(opens in a new tab)
Moderate evidence599 students assessed twice, of whom 100 broke up
A prospective study measuring students before and after a breakup, against a comparison group who did not break up, found high reported growth but concluded the pattern was better explained by positive reappraisal than by demonstrated positive change.
What this does not show
A ten week interval is short for detecting genuine developmental change, and the sample is again college students in relatively brief relationships. A null result over ten weeks does not establish that no growth occurs across longer periods.
Owenz, M., & Fowers, B. J., Perceived post-traumatic growth may not reflect actual positive change: A short-term prospective study of relationship dissolution (2019)(opens in a new tab)
Moderate evidence2,484 divorcing parents, including 329 matched couples
Among parents who had already filed for divorce and attended a mandatory parenting class, around one in four individuals still believed the marriage could be saved, and in roughly one in nine matched couples both partners did. Interest in reconciliation was considerably higher among those whose spouse had initiated the divorce.
What this does not show
A single-state American sample of divorcing parents surveyed at one moment, measuring belief and interest rather than regret, with no follow-up establishing what those people concluded later or whether any reconciled.
Doherty, W. J., Willoughby, B. J., & Peterson, B., Interest in marital reconciliation among divorcing parents (2011)(opens in a new tab)

How we rate evidence strength

Strong evidence: Multiple independent studies, including replications or meta-analyses, point the same way.
Moderate evidence: Several studies support this, but samples are limited or findings vary by population.
Limited evidence: Early, small, or single-study evidence. Treat this as a reasonable hypothesis, not a settled fact.
Contested: Credible researchers disagree, or the evidence points in conflicting directions.

Comparative view

Different perspectives

Traditions and disciplines answer this differently. We describe what each one teaches. We are not telling you which is right, and no tradition speaks with a single voice.

What the Evidence Supports

The strongest claim available here is the affective forecasting one, not because the breakup study is large, since it analysed twenty six people, but because it sits inside a much wider literature on emotional misprediction that has replicated broadly. The cycling findings are well measured and consistent across dating, cohabiting and married samples, but they are correlational throughout and cannot say whether cycling harms relationships or merely marks ones that were already unstable. The life satisfaction result is the only item here drawn from a national sample across a decade, and it is also the most often misreported, since it applied to women leaving very low quality marriages rather than to divorcing people in general. On regret itself, measured directly, there is essentially nothing, and the honest position is that the question has not been studied properly rather than that the answer is known.

Therapist Perspective

Clinicians working with post-breakup distress commonly describe the first task as slowing the interpretation down rather than resolving it. A pattern practitioners report often is the client who arrives certain they made a catastrophic mistake, and whose account of the relationship has quietly reorganised itself around the good parts, which is a documented feature of how memory behaves under loss rather than a signal of buried truth. Therapists also tend to separate missing a particular person, missing being partnered, and missing the version of the future that has now gone, on the grounds that these respond to quite different things and that lumping them together produces a vague ache with nothing to act on. Where contact with the ex is continuing, most will raise it early, since intermittent contact tends to restart the clock on a process that was already underway.

Secular Perspective

A secular reading treats regret as information about you rather than as a verdict handed down from outside, and declines to grant it authority simply because it is painful. On this view the relevant question is not whether you feel regret but whether the reasons you acted on were sound at the time and remain sound now that you know more. It also puts weight on what is owed to the other person, who is not a variable in your recovery: reopening contact to relieve your own doubt, with no intention of actually returning, counts as a real wrong rather than a minor one. And it holds that a decision made carefully on the information available was a good decision even where the outcome turned out badly, a distinction people abandon very quickly when they are hurting.

Practical steps

What you can actually do

Write your reasons down now, dated, in your own words, while you still have direct access to them. The reasons erode first, and reconstructing them from memory in three months is not the same as reading them back.
Treat the first month's certainty in either direction as low quality information, since that is where the forecasting error is concentrated. This cuts both ways: early relief is not a verdict any more than early anguish is.
If you are considering going back, name the specific thing that ended it and ask what has actually changed about that thing. The cycling research suggests reunions where nothing structural changed tend to reproduce the original problem with more uncertainty attached.
Keep the grief and the verdict in separate sentences. 'I miss him' and 'I was wrong' are different claims, and running them together makes both harder to examine.
Hold the lessons-learned narrative loosely in the early months. It is a useful way to live with what happened and a poor instrument for deciding whether the decision was right.

Worth unlearning

Common misconceptions

That wanting your ex back means you made a mistake. The urge is close to universal, peaks early, and is strongest at the point when predictions about your own feelings are least accurate.
That there is a known percentage of people who regret breaking up. No study has measured it in a representative sample, and the figures in circulation do not come from research.
That getting back together resolves it. Reunions are very common and the renewed relationships rate lower on satisfaction and higher on uncertainty than relationships that never split.
That everyone feels better within a year. The study tracking life satisfaction across divorce found the gain among women leaving very low quality marriages, and not among those leaving better ones.
That feeling you have grown proves you were right. A prospective study with a comparison group found reported growth after a breakup was better explained by reappraisal than by measured change.

The short version

Key takeaways

No representative study has measured how many people regret a breakup, so any percentage in circulation is not a research finding.
Research
Forecasts of post-breakup distress run systematically too pessimistic, and the error sits in the predicted initial intensity rather than in the predicted rate of recovery.
Research
Breaking up and reconciling is common, and renewed relationships report on average fewer positives and more uncertainty than relationships that never split.
Research
Gains in life satisfaction after divorce were found among women leaving very low quality marriages, not across divorcing people generally.
Research
Grief over a loss and the judgement that a decision was wrong are different things that arrive together and are routinely mistaken for one another.
Clinical consensus

Worth saying

When to get professional help

Ordinary breakup distress is intense and does get less so. Distress that has not shifted at all after a few months, or that is stopping you sleeping, eating or working, is worth taking to a doctor or a therapist rather than waiting out, and the same applies if you are drinking more in order to manage it. Thoughts of harming yourself are a reason to contact a crisis line or emergency service the same day rather than a reason to wait for an appointment. If you are considering reconciling with someone you left because of violence, threats or a pattern of control, the period around returning and around leaving again carries elevated risk, and a domestic abuse service is the right place for that conversation rather than a couples counsellor. Where an ex will not accept the end and contact is escalating or frightening, that is not a reconciliation question and should be treated as its own problem. For the ordinary case, individual therapy tends to be more useful than anything joint, since the work is on your own account of what happened.

Find support services near you

Still wondering

What other people say

These are readers describing their own relationships, not professionals. Take them as experience, not advice, and please do not name anyone.

No one has shared their experience yet. If you have been here, you would be the first, and probably the most useful thing on this page.

Check our work

Sources

[1]
Mispredicting distress following romantic breakup: Revealing the time course of the affective forecasting error(opens in a new tab)
Eastwick, P. W., Finkel, E. J., Krishnamurti, T., & Loewenstein, G. · Journal of Experimental Social Psychology · 2008
[2]
On-again/off-again dating relationships: How are they different from other dating relationships?(opens in a new tab)
Dailey, R. M., Pfiester, A., Jin, B., Beck, G., & Clark, G. · Personal Relationships · 2009
[3]
"It's complicated": The continuity and correlates of cycling in cohabiting and marital relationships(opens in a new tab)
Vennum, A., Lindstrom, R., Monk, J. K., & Adams, R. · Journal of Social and Personal Relationships · 2014
[4]
Women in very low quality marriages gain life satisfaction following divorce(opens in a new tab)
Bourassa, K. J., Sbarra, D. A., & Whisman, M. A. · Journal of Family Psychology · 2015
[5]
"I'll never be in a relationship like that again": Personal growth following romantic relationship breakups(opens in a new tab)
Tashiro, T., & Frazier, P. · Personal Relationships · 2003
[6]
Perceived post-traumatic growth may not reflect actual positive change: A short-term prospective study of relationship dissolution(opens in a new tab)
Owenz, M., & Fowers, B. J. · Journal of Social and Personal Relationships · 2019
[7]
Interest in marital reconciliation among divorcing parents(opens in a new tab)
Doherty, W. J., Willoughby, B. J., & Peterson, B. · Family Court Review · 2011
[8]
Immune neglect: A source of durability bias in affective forecasting(opens in a new tab)
Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. P. · Journal of Personality and Social Psychology · 1998

breakups getting back together regret reconciliation moving on

People who read this also asked

Breakups
How do you know when to end a relationship?
11 min read