What Makes a Marriage Last? What Long-Term Research Shows
Relationship Health · · Updated · 10 min read
Short answer
Long-term research suggests marriages last when partners stay committed, keep appreciation and affection alive, believe each other is responsive, and handle conflict without letting negativity take over. In long studies, a slow loss of love and affection marked couples headed for divorce more than rising conflict did. Personality predicted relationship quality less well than feelings about the relationship.
By the Compatify Editorial Team · Sources checked . Science-informed self-reflection, not clinical advice. How we write and check guides.
At a golden wedding anniversary, someone always asks the question. The couple have been married fifty years, the grandchildren are restless, and a cousin raises a glass and says: so what is the secret? He says never go to bed angry. She laughs and says separate bathrooms. Everyone claps, and everyone quietly wonders whether either answer is true, and whether their own relationship has whatever it is.
Advice from long-married couples is warm and often wise, but it is one couple's story. Researchers have tried to answer the same question with something sturdier, following couples from the early years of marriage to see which ones stay together and stay happy. Their answers are less quotable than a toast, and more useful.
What makes a marriage last?
What makes a marriage last, according to long-term studies, is a combination of commitment that both partners can feel, warmth and affection that keep being expressed, a belief that each partner is responsive to the other, and conflict that does not tip into sustained negativity, all while the couple copes with whatever stress life brings. The research describes patterns across many couples, not a formula for yours. Much of this research reaches beyond marriage: Joel and colleagues' analysis covered romantic couples generally, and Spanier designed the Dyadic Adjustment Scale for married and unmarried cohabiting couples alike.
What does long-term research actually show?
Long-term research shows that how a marriage changes over time depends on what partners bring, what happens to them, and how they handle it together. Benjamin Karney and Thomas Bradbury reviewed 115 longitudinal studies, representing over 45,000 marriages, in 1995. From that review they outlined an integrated model, now usually called the vulnerability-stress-adaptation model. The model proposes that each partner's enduring vulnerabilities, such as personality or family background, combine with the stresses the couple faces, and that their adaptive processes, how they talk, solve problems and support each other, shape whether the marriage grows stronger or weaker.
| Factor | What the research suggests | Where it comes from |
|---|---|---|
| Feeling your partner is committed | One of the strongest predictors of relationship quality | Joel and colleagues, 2020, 43 longitudinal studies |
| Appreciation | Among the top predictors of relationship quality | Joel and colleagues, 2020 |
| Love and affection that keep being expressed | Their gradual loss marked couples headed for divorce | Huston and colleagues, 2001, 13-year follow-up |
| Positive emotion in everyday conversation | Its absence predicted divorces that came later in marriage | Gottman and Levenson, 2000, 14-year study |
| Negativity during conflict | Predicted divorces that came early in marriage | Gottman and Levenson, 2000 |
| Stress the couple faces | Shapes the course of a marriage through how partners adapt to it | Karney and Bradbury, 1995, review of 115 studies |
| Personality and other individual traits | Predicted some quality, but added nothing once feelings about the relationship were counted | Joel and colleagues, 2020 |
Is it love that fades, or fighting that ends a marriage?
Both can, and two long American studies suggest that fading warmth deserves as much attention as fighting. John Gottman and Robert Levenson followed couples for 14 years and found that negative emotion during conflict predicted divorces that came early, while a lack of positive emotion, in everyday conversations and in arguments, predicted divorces that came later. Ted Huston and his colleagues followed newlyweds for 13 years and found little support for the idea that rising negativity early in marriage leads to failure. Instead, disillusionment distinguished couples headed for divorce: love fading, open affection declining, a weakening belief that the spouse was responsive, and growing ambivalence. Avoiding fights is only half the work; the other half is keeping warmth alive on ordinary days. The four horsemen guide covers the conflict side, and the guide to partner responsiveness covers feeling understood.
What makes a marriage last 50 years?
Nobody can say with confidence what makes a marriage last 50 years, because very few studies follow couples for that long. The longest studies in this guide ran for 13 and 14 years. What they suggest is that the patterns of the early years carry forward: in Huston's study, the intensity of newlyweds' romance and how much negative feeling they expressed predicted both whether couples were happy 13 years later and how long the marriages that ended had lasted. It is reasonable to guess that couples married half a century kept doing what these studies link to staying together, such as feeling committed and appreciated. That is an inference from shorter studies, not a finding about 50-year marriages, and those couples were also lucky with health and circumstance.
Does personality compatibility make a marriage last?
Personality compatibility appears to matter less than people's day-to-day feelings about the relationship itself. In 2020 Samantha Joel and a large team of collaborators used machine learning across 43 longitudinal couples datasets from 29 laboratories. The top relationship-specific predictors of relationship quality were perceived partner commitment, appreciation, sexual satisfaction, perceived partner satisfaction and conflict. Individual differences, such as life satisfaction, negative mood, depression and attachment anxiety or avoidance, did predict quality too, but they had no predictive effect beyond a person's own view of the relationship. All the datasets came from what the authors call Western countries: the United States, Canada, Switzerland, New Zealand, the Netherlands and Israel. The authors suggest that individual differences seem to exert their influence through how people experience the relationship. That is relationship quality, not divorce, but it points the same way: what happens between you counts for more than who each of you is on paper.
What are the secrets of a long marriage?
The secrets of a long marriage, as far as research can tell, are habits. These draw on the patterns the studies above link to lasting relationships:
- Say out loud that you are in it for the long haul; feeling your partner's commitment counts.
- Thank each other for ordinary things, like cooking, driving or remembering the bins, and mean it.
- Keep affection going on days when nothing special is happening: a touch on the way past, a real hello at the door.
- Ask about each other's day and listen, since everyday talk is where warmth can quietly fade.
- Complain about behaviour, not character, and repair quickly when an argument goes wrong.
- Treat stress as something you face together, and name it when a hard season is spilling into the relationship.
- Talk about intimacy and sex as a normal part of the relationship, not an emergency topic.
- Check in on your shared goals every so often, so you are still building the same life.
- Notice ambivalence early. If you feel yourself pulling back, say so before it hardens into distance.
- Consider counselling before things reach a crisis, not only at the end.
Shared goals and rituals are a large part of why some couples keep feeling like a team, and the guide to shared meaning explains how couples build them.
Can you predict whether a marriage will last?
You cannot reliably predict whether a particular marriage will last, even with good research. Gottman and Levenson's 2000 paper reported a model that predicted divorce with 93 per cent accuracy, but that figure describes how well a model fitted the same 79 couples it was built on, and when Richard Heyman and Amy Slep tested divorce prediction equations on new couples in 2001, a step called cross-validation, accuracy fell sharply. Joel and colleagues found that relationship-specific variables predicted up to 45 per cent of the variance in relationship quality at the start of the studies, and that changes in quality over time, people becoming happier or less happy, were largely unpredictable from any combination of self-report measures.
Where does the science of lasting marriages come from?
The science of lasting marriages grew from attempts to measure marital quality in the middle of the twentieth century. Harvey Locke and Karl Wallace published the Marital Adjustment Test in 1959, and Graham Spanier published the Dyadic Adjustment Scale in 1976, which was designed for married and unmarried cohabiting couples alike. From the early 1980s, American researchers such as Gottman and Levenson, and Huston's PAIR Project, began following couples for years, and Karney and Bradbury's 1995 review pulled 115 longitudinal studies, those included, together. In 2007 Janette Funk and Ronald Rogge used item response theory to build the Couples Satisfaction Index, showing that it measured satisfaction more precisely than the older scales. The 2020 project led by Joel brought dozens of research groups together to ask which predictors held up across all their data.
How do you check how satisfied you are in your own relationship?
Compatify's Relationship Satisfaction Test is designed to help you reflect on how fulfilled you feel in your relationship, beyond surface happiness. It looks at conflict resolution, emotional support, shared goals, intimacy and trust, the kinds of areas long-term research ties to relationships that last. Your first test is free, and your result shows your overall score and what it means, with one strength and one growth area from your answers. If your partner takes it as well, you can compare your result with theirs. For a present-day check, see the guide to knowing if your relationship is healthy.
The framework behind this
Karney and Bradbury's 1995 review in Psychological Bulletin summarised 115 longitudinal studies and outlined the model linking vulnerabilities, stress and adaptation. Gottman and Levenson's 2000 paper in the Journal of Marriage and Family separated the early and late paths to divorce. Huston and colleagues' 2001 study in the Journal of Personality and Social Psychology showed that disillusionment, more than rising conflict, marked marriages that ended. Joel and colleagues' 2020 paper in the Proceedings of the National Academy of Sciences identified the predictors that held up across 43 studies. Funk and Rogge's 2007 paper in the Journal of Family Psychology developed the Couples Satisfaction Index and compared it with the Dyadic Adjustment Scale. Compatify's test is grounded in the Couples Satisfaction Index (CSI-32) and the Dyadic Adjustment Scale. It is science-informed self-reflection, not a clinical evaluation.
What this cannot tell you
A result describes how satisfied you feel right now, as you report it. It cannot tell you whether your marriage will last or what your partner feels. The studies in this guide describe averages: Gottman and Levenson's and Huston's couples were American, and Joel's datasets all came from Western countries, so a pattern across those couples is not a prediction for yours. A long marriage is also not the goal in itself. If a relationship involves violence, threats, control over money or friends, or fear of your partner, safety comes first, and leaving is best planned with support, because leaving can raise the risk. In Australia, 1800RESPECT is on 1800 737 732 at any hour, and anyone in immediate danger should call 000. If you are struggling, Lifeline is on 13 11 14.
Common questions
- What is the number one thing that makes a marriage last?
- No single factor decides it, but long-term studies keep returning to commitment and warmth. Joel and colleagues found perceived partner commitment and appreciation among the strongest predictors of relationship quality, and Huston's study linked fading love and affection to divorce.
- What makes a marriage last long?
- Research points to feeling that your partner is committed, keeping appreciation and affection going on ordinary days, believing your partner is responsive to you, handling conflict without letting negativity take over, and adapting together when stress arrives.
- Do couples who argue less stay together longer?
- Not necessarily. In Gottman and Levenson's 14-year study, negativity in conflict predicted early divorce, but a lack of positive emotion predicted later divorce. Couples who never argue can still drift apart if warmth fades between them.
- Does having similar personalities make a marriage last?
- The evidence suggests it matters less than expected. In Joel and colleagues' study of 43 datasets, personality and other individual differences added no predictive power once people's own feelings about the relationship were taken into account.
- Is it normal for marriage satisfaction to go up and down?
- Yes. Satisfaction can change over the years, and Joel and colleagues found those changes largely unpredictable from self-report measures. A low stretch deserves attention, but it is not a verdict on the marriage.
Sources
- Heyman, R. E. and Slep, A. M. S. (2001). The hazards of predicting divorce without crossvalidation. Journal of Marriage and Family.
- Karney, B. R. and Bradbury, T. N. (1995). The longitudinal course of marital quality and stability: a review of theory, method, and research. Psychological Bulletin.
- Gottman, J. M. and Levenson, R. W. (2000). The timing of divorce: predicting when a couple will divorce over a 14-year period. Journal of Marriage and Family.
- Huston, T. L., Caughlin, J. P., Houts, R. M., Smith, S. E. and George, L. J. (2001). The connubial crucible: newlywed years as predictors of marital delight, distress, and divorce. Journal of Personality and Social Psychology.
- Joel, S., Eastwick, P. W., Allison, C. J. and colleagues (2020). Machine learning uncovers the most robust self-report predictors of relationship quality across 43 longitudinal couples studies. Proceedings of the National Academy of Sciences.
- Funk, J. L. and Rogge, R. D. (2007). Testing the ruler with item response theory: increasing precision of measurement for relationship satisfaction with the Couples Satisfaction Index. Journal of Family Psychology.