Do Opposites Attract? Similarity vs Complementarity in Psychology

Do Opposites Attract? The Short Answer

More precisely, a meta-analysis of 313 studies (Montoya, Horton & Kirchner, 2008) found that both actual and perceived similarity are strongly related to attraction, and that in established relationships it is perceived similarity, the sense that "we are alike where it counts," that keeps predicting attraction. No comparable body of evidence exists for complementarity.

The rest of this article walks through where each hypothesis came from, what the research record actually shows, and how to apply the findings when choosing or understanding a partner.

Two Competing Hypotheses

In love and human relationships, "birds of a feather flock together (similarity hypothesis)" and "opposites attract (complementarity hypothesis)" are among the oldest debated topics. Since everyday intuition can summon examples of both, many people find it hard to judge which is correct.

However, the accumulation of psychological research provides a clear answer. Bottom line: in long-term relationships, similarity has an overwhelming advantage. While "opposites attract" is partially correct at the early attraction stage, similarity consistently dominates in relationship maintenance and satisfaction.

Research Supporting the Similarity Hypothesis

Since Byrne (1971)'s "similarity-attraction paradigm," hundreds of studies have reported positive correlations between attitude, value, and personality similarity and interpersonal attraction.

Montoya, Horton & Kirchner (2008)'s meta-analysis integrated 460 effect sizes from 313 laboratory and field studies and found large positive associations with attraction: r = .47 for actual similarity and r = .39 for perceived similarity (the degree to which people believe they are alike). The same analysis added an important nuance: the effect of actual similarity was strong in short laboratory encounters but was not significant in existing relationships, whereas perceived similarity predicted attraction in laboratory studies and existing relationships alike. In real couples, feeling similar matters even more than objectively being similar.

For personality traits, matching is often confused with a related finding: Malouff et al. (2010)'s meta-analysis examined the link between one partner's Big Five trait levels and the other partner's relationship satisfaction. In other words, what your partner is like carries weight in its own right, alongside how well the two of you match.

Several mechanisms explain why similarity is advantageous. (1) Self-validation: similar partners give the sense of affirming one's values. (2) Predictability: similar partners' behavior is easier to predict, generating reassurance. (3) Shared activities: when interests and values align, more activities can be enjoyed together. (4) Reduced conflict: when fundamental values align, daily decision-making conflicts are fewer.

Limited Validity of the Complementarity Hypothesis

The complementarity hypothesis ("opposites attract") isn't entirely refuted. However, its scope of application is very limited.

Winch (1958)'s proposed complementarity hypothesis assumed specific complementary interpersonal needs: "dominant people are drawn to submissive people," "caregivers are drawn to those needing care." But subsequent replication research found limited support for this hypothesis.

Modern research shows complementarity functions mainly in these limited contexts. (1) Early attraction stage: feeling freshness or admiration toward partners with traits one lacks. (2) Specific role divisions: when functional division of labor works (one extraverted handling socializing, the other introverted maintaining the home). (3) Short-term relationships: in short-term relationships where stimulation and novelty matter, dissimilarity can be attractive.

However, in long-term relationships, the initial "freshness" fades and daily value clashes come to the foreground. "Interesting because we're opposites" easily becomes "exhausting because we're opposites" within months.

Optimal Compatibility - Similarity + Moderate Difference

The research record does not demand "complete similarity." A practical reading of the evidence is "high similarity where it counts, plus room for difference."

Gonzaga, Carter & Buckwalter (2010) studied married couples who had met through an online matching service, examining spouses' personality similarity (assortative mating), their convergence over time, and how these relate to marital satisfaction (Personal Relationships, 2010). Meanwhile, the meta-analytic finding that perceived similarity outweighs actual similarity in existing relationships (Montoya et al., 2008) implies that a couple does not need to match on everything: what sustains attraction is the shared sense of being alike on the things both partners consider central.

This is intuitively understandable. Two completely identical people cannot offer each other new perspectives. Moderate difference brings stimulation and growth opportunities to the relationship. What matters is "where to be similar and where to differ."

Areas to align: values, life goals, conscientiousness, agreeableness, emotional stability

Areas where difference is acceptable: hobbies, level of extraversion, openness (attitude toward new experiences)

This site's compatibility assessment reflects this insight, evaluating similarity on all traits while assigning higher weights to similarity in agreeableness, conscientiousness, and emotional stability.

Practical Advice

Specific advice for applying scientific findings to daily life.

Partner selection stage: Beyond initial "butterflies" and "freshness," prioritize alignment in values and lifestyle. "Can we naturally agree about daily meals, weekend activities, and how to spend money?" is an important indicator predicting long-term relationship quality.

Improving existing relationships: If you already have major differences with your partner, it's important to "understand" rather than "change" them. Acknowledging the existence of differences and respecting each other's domains - this "agreed-upon difference" doesn't threaten relationships. What threatens them is the pressure to "make the partner conform to oneself."

Importance of self-understanding: Before considering compatibility, accurately understanding your own personality traits is crucial. Knowing what you value and in what environments you feel most comfortable, the characteristics of compatible partners naturally become visible.

Frequently Asked Questions

Q. Do opposites really attract?
Only in limited contexts. Dissimilarity can feel fresh at the early attraction stage, and some couples make functional role divisions work. Across the 313 studies in Montoya et al. (2008)'s meta-analysis, however, it is similarity, and especially perceived similarity, that predicts attraction, and long-term satisfaction favors couples aligned on core values.

Q. What is the complementarity hypothesis?
Proposed by sociologist Robert Winch in 1958, it holds that people choose partners whose interpersonal needs complement their own, such as a dominant person pairing with a submissive one. Later replication studies found only limited support, so it is treated as applying to narrow contexts (early attraction, specific role divisions) rather than as a general law of attraction.

Q. My partner and I are very different. Is that a bad sign?
Not by itself. What matters most is alignment on core values and life goals, plus the shared sense of being alike where it counts. Differences in hobbies, energy levels, or openness to new experiences can even benefit a relationship when both partners acknowledge and respect them instead of trying to change each other.