Do opposites really attract?

Largely no. It is one of the most repeated claims about relationships and one of the least supported — the research consistently points the other way.

What the evidence shows. Decades of work on assortative mating find that partners resemble each other on a wide range of characteristics: age, education, political and religious views, intelligence, attitudes and values, socioeconomic background, and even height and physical attractiveness. A large 2023 analysis covering many traits across millions of couples found positive correlations for the substantial majority of traits examined and very few negative ones. Similarity is the rule; complementarity is the exception.

Why similarity works:

Proximity produces it before anything else does. You meet people through school, work, neighbourhood and social circles, all of which are already sorted by background and interests. A large share of the effect is simply who you encounter.

Shared attitudes reduce friction. Agreeing on money, children, religion, politics and how to spend time removes whole categories of recurring conflict.

Validation. Encountering someone who sees the world as you do is intrinsically rewarding, and research on attitude similarity and attraction is among the more robust findings in social psychology.

Where difference genuinely helps. Complementarity appears to work for traits and roles rather than values — one person more organised and the other more spontaneous, or differing dominance in particular domains. Personality similarity, notably, predicts relationship satisfaction less strongly than attitude similarity does.

Why the myth persists. Differences are noticeable and story-worthy in a way similarities are not; nobody remarks that a couple share a politics. Early attraction can also be heightened by novelty, which fades — and the same difference that was intriguing at three months is often the perpetual argument at three years.