Is it normal to fight a lot in a healthy relationship?
The short answer
Frequency on its own turns out to be a weak signal. Longitudinal research finds that couples who argue often can be as stable as couples who rarely do, and that what separates them is what happens during the argument rather than how many there are. Contempt, stonewalling, defensiveness and criticism are the behaviours associated with poor outcomes. A couple that argues weekly and repairs well is in better shape than one that argues twice a year and does it badly.
What readers think
When you argue, do you usually find your way back to each other the same day?
Anonymous. One vote per reader.
The worry usually arrives in a comparison. Someone knows a couple who never seem to argue, and measures their own household against that, and concludes something has gone wrong. The research on couples followed over years points somewhere else entirely, and the place it points is more demanding than a frequency count but also considerably more hopeful, because it describes something a couple can actually change.
Frequency is a poor predictor on its own
Observational studies that recorded couples arguing in a laboratory and then followed them for years consistently find that the amount of conflict is not what separates the couples who stay together from those who do not[1]. Some stable couples argue rarely and quietly. Others argue often and loudly and remain stable for decades. What the recordings distinguish is the character of the exchange: whether disagreement stays about the issue, whether either person expresses disdain for the other, and whether attempts to calm things down succeed. This is a genuinely different question from how often the arguments happen, and it explains why the couple who never argue may not be doing better. Silence can indicate a settled partnership, or it can indicate that one person has stopped raising things because raising them has never once gone well.
The four behaviours that carry the signal
The behaviours identified as most corrosive during conflict are criticism, which attacks the person rather than the problem, defensiveness, which meets a complaint by reversing it, stonewalling, which withdraws from the conversation entirely, and contempt, which communicates that the other person is beneath you through sarcasm, mockery, eye-rolling or name-calling[2]. Contempt is treated as the most damaging of the four in this literature, and it is the one most reliably absent from couples who do well. The practical value of this framing is that these are behaviours rather than personality traits. A person who rolls their eyes during arguments is doing something, and doing something is easier to stop than being something. Couples who reduce these specific behaviours tend to see the arguments themselves become less frightening even when the underlying disagreements remain unresolved.
Repair matters more than resolution
A finding that surprises people is how many of a couple's disagreements are never actually settled. Long-term observational work suggests that a substantial majority of the issues couples argue about are perpetual, rooted in stable differences in personality, upbringing or preference, and that these same arguments recur across decades. If resolution were the standard, almost every marriage would be failing. What distinguishes the couples who do well is repair: the joke that breaks the tension, the hand on the arm, the sentence that concedes a point, the return to the conversation an hour later in a better register. Research following couples across sixteen years found that patterns of constructive versus destructive conflict behaviour, and changes in those patterns over time, were associated with whether the marriage lasted[3]. The recurring argument is not the problem. The recurring argument with no way back out of it is.
What the predictions can and cannot do
This literature is often reported with striking accuracy figures attached, and those figures deserve scrutiny. Statistical critiques have pointed out that several widely cited divorce prediction models were fitted and evaluated on the same sample, which inflates apparent accuracy, and that the models performed considerably less well when applied to couples they had not been built on[4]. This matters for how you should read any claim that a researcher can watch a couple for fifteen minutes and predict their divorce. The underlying observation that certain conflict behaviours are associated with poorer outcomes has held up across independent samples. The claim that any individual couple's future can be read off a short recording has not, and treating it as though it can is a misuse of the research.
It depends, on what?
What changes the answer
Two households can produce identical argument counts and be in completely different situations, which is why the number tells you so little. There is also a cultural layer that rarely gets mentioned in the research, most of which was conducted with North American couples. Volume, interruption and direct disagreement carry different meanings across families and cultures, and an exchange that reads as alarming in one setting is ordinary conversation in another. What travels better than the volume is the presence or absence of contempt, which tends to be recognised as an insult wherever it occurs. Separately, it is worth naming that none of this applies to a relationship where one person is frightened. Conflict research describes disagreements between two people who are broadly equal in the exchange, and it is not a framework for understanding intimidation, coercion or violence, which are different in kind and not in degree.
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
- Frequent arguing often indicates that both people still consider the relationship worth contesting things in, which is not true of every quiet household.
- Long-term observational research finds stable couples across a wide range of conflict frequencies, so no single rate is diagnostic.
- Most disagreements between committed partners are perpetual rather than solvable, which makes recurrence expected rather than evidence of failure.
- Focusing on behaviours during conflict gives a couple something specific to change, whereas focusing on frequency does not.
The case for no
- High conflict is genuinely costly even when a relationship survives it, and the strain on children in the household is well documented independently of whether the parents separate.
- Some couples use frequency as reassurance to avoid examining an argument style that has become cruel, since the research is regularly quoted as though all fighting is healthy.
- Escalating frequency over time is a meaningful change even though the absolute number is not, and a couple arguing far more than they used to has information worth taking seriously.
- The research base is heavily weighted toward Western, mostly white, mostly married samples, which limits how confidently it can be generalised.
Evidence
What the research says
- Moderate evidenceDozens of couples per cohort, followed for four to fourteen years
Couples observed during conflict and followed for years showed that specific interaction behaviours, rather than the amount of conflict, distinguished those who separated from those who remained together.
Gottman, J. M., & Levenson, R. W., Marital processes predictive of later dissolution: Behavior, physiology, and health (1992)(opens in a new tab)What this does not show
Based on laboratory conversations that may not represent how couples argue at home, with samples drawn largely from white American married couples in specific regions and decades. Effect sizes are for groups and do not predict individual outcomes.
- Moderate evidenceSmall longitudinal cohorts observed repeatedly over years
Criticism, defensiveness, stonewalling and contempt during conflict were each associated with marital instability, with contempt showing the strongest association.
Gottman, J. M., & Levenson, R. W., The timing of divorce: Predicting when a couple will divorce over a 14-year period (2000)(opens in a new tab)What this does not show
Coded from observed laboratory interaction by trained raters, which is more reliable than self-report but still interpretive. The four behaviours co-occur, making their individual contributions hard to separate.
- Moderate evidenceAround 370 couples followed for sixteen years
A sixteen-year study of married couples found that conflict behaviour patterns, and how they changed over time, were associated with the likelihood of divorce, with the associations differing by gender and by race.
Birditt, K. S., Brown, E., Orbuch, T. L., & McIlvane, J. M., Marital conflict behaviors and implications for divorce over 16 years (2010)(opens in a new tab)What this does not show
Relies on self-reported conflict behaviour rather than observation. Drawn from one American longitudinal sample of couples marrying in a single year, so period effects cannot be ruled out.
- Strong evidenceRe-analysis of published prediction studies
A methodological review found that prominent divorce prediction models had been evaluated on the same data used to build them, and that reported accuracy fell substantially under proper cross-validation.
Heyman, R. E., & Slep, A. M. S., The hazards of predicting divorce without crossvalidation (2001)(opens in a new tab)What this does not show
A critique of method rather than new data on couples. It challenges the precision of the predictions, not the existence of the underlying associations.
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 defensible claim from this literature is associational: certain observed behaviours during conflict correlate with later separation across multiple samples. The indefensible claim is predictive precision about a specific couple, which the cross-validation critique undermines directly. It is also worth noting that most of this work is correlational, so the possibility that contempt is a symptom of a relationship already failing, rather than a cause of its failure, is not fully excluded by the designs used. The behaviours are worth changing either way, but the causal story is less settled than popular accounts suggest.
Therapist Perspective
Couples therapists commonly report that the frequency question is not the one worth answering, and that the more useful line of enquiry is what each argument is actually about underneath the stated topic. The recurring argument about dishes is rarely about dishes; it is usually about whether one person feels their effort is seen. Clinicians also tend to distinguish sharply between couples who fight hard and reconnect afterwards and couples who fight quietly and stay disconnected for days, and they generally treat the second as the harder problem despite it looking calmer from outside. This reflects widespread clinical practice rather than a finding from controlled trials.
Secular Perspective
A secular reading treats conflict as the ordinary cost of two separate people sharing a life, and locates the ethical question not in whether they disagree but in how they treat each other while disagreeing. On this view there is nothing inherently virtuous about a peaceful household and nothing inherently wrong with a noisy one. What matters is whether both people retain standing in the argument, whether disagreement is permitted without penalty, and whether the terms of the relationship are open to renegotiation rather than fixed by whoever is more willing to escalate.
Practical steps
What you can actually do
Watch for contempt specifically rather than counting arguments. Sarcasm, mockery and eye-rolling are the behaviours most consistently associated with poor outcomes, and noticing them is the first step to dropping them.
Learn the difference between withdrawing to calm down and withdrawing to punish. Saying you need twenty minutes and naming when you will return is a repair; leaving the room without a word is not.
Accept that some arguments will recur permanently and aim for a better version of the same argument rather than a final resolution.
Complain about the behaviour rather than the person. The difference between saying a task was not done and saying someone is lazy is the difference between a complaint and an attack.
Notice whether a quiet period is peace or avoidance. If either of you has a list of things you no longer bother raising, the absence of arguing is a symptom rather than a result.
Worth unlearning
Common misconceptions
That healthy couples do not argue. Observational research finds stable couples across a wide range of conflict frequencies, including very high ones.
That every argument should end in resolution. A large share of the issues couples argue about are perpetual and recur for the life of the relationship.
That researchers can predict divorce from a short recording with the accuracy often quoted. Those figures come largely from models tested on the same couples they were built on, and they perform worse on new samples.
That raising your voice is the danger sign. Volume is culturally variable and weakly predictive; contempt is neither.
That a couple who never argue must be doing well. Absence of conflict can equally mean one person has given up on being heard.
The short version
Key takeaways
How often a couple argues is a weak predictor of whether the relationship lasts.
ResearchCriticism, defensiveness, stonewalling and contempt during conflict are associated with poorer outcomes, with contempt the most damaging.
ResearchMost disagreements between long-term partners are perpetual rather than solvable, so repair matters more than resolution.
ResearchWidely quoted divorce prediction accuracy figures are inflated by testing models on the samples that produced them.
ResearchConflict research does not describe relationships involving fear, intimidation or violence, which require a different response.
Clinical consensus
Worth saying
When to get professional help
If any argument has involved physical force, threats, or one person being made afraid, that is outside what conflict research describes and outside what couples counselling is designed for; a domestic abuse service is the appropriate first contact, and joint counselling is generally advised against in that situation because it can increase risk. Short of that, a couples therapist is worth considering when the same argument has run for years with no repair available, when one person consistently shuts down and cannot re-enter the conversation, or when contempt has become the normal register. Individual therapy can be more useful than couples work when the intensity of your own reactions during conflict feels disproportionate to what is happening, which often has more to do with your own history than with your partner.
Still wondering
Related questions people ask
There is no threshold in the research that marks too much. Daily arguing that stays about the issue and ends in reconnection is different from daily arguing that involves contempt or leaves both people isolated for hours. The more useful question is what the arguments look like and how they end.
Community
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]Marital processes predictive of later dissolution: Behavior, physiology, and health(opens in a new tab)
Gottman, J. M., & Levenson, R. W. · Journal of Personality and Social Psychology · 1992
- [2]The timing of divorce: Predicting when a couple will divorce over a 14-year period(opens in a new tab)
Gottman, J. M., & Levenson, R. W. · Journal of Marriage and Family · 2000
- [3]Marital conflict behaviors and implications for divorce over 16 years(opens in a new tab)
Birditt, K. S., Brown, E., Orbuch, T. L., & McIlvane, J. M. · Journal of Marriage and Family · 2010
- [4]The hazards of predicting divorce without crossvalidation(opens in a new tab)
Heyman, R. E., & Slep, A. M. S. · Journal of Marriage and Family · 2001