Cohabitation is sometimes thought of as being inversely associated with education, but in Britain a more complex picture emerges. Educational group differences in cohabitation vary by age, time period, cohort, and indicator used. Well‐educated women pioneered cohabitation in Britain in the 1970s and 1980s. In the most recent cohorts, however, the less educated have exceeded the best educated in the proportions ever having cohabited at young ages. But the main difference by education currently seems largely a matter of timing—that is, the less educated start cohabiting earlier than the best educated. In Britain, educational differentials in cohabitation appear to be reinstating longstanding social patterns in the level and timing of marriage. Taking partnerships as a whole, social differentials have been fairly stable. Following a period of innovation and diffusion, there is much continuity with the past.