This paper is focused on couple households where the wife is the main earner. The economic literature on this subject is particularly scant. According to our estimates, the wife was the main earner in one of every six couple households in France in 2002, including wife-sole-earner households. The proportion of wives outearning their husbands was 18% for dual-earners. About 24% of American women in dual-earner households earned more than their husband in 2004. Using a model of household labour supply behaviour, we show that households where the wife is the main earner may come about either because the husband has a weaker preference for work than his wife, due possibly to her high wage, or because he is hit by adverse circumstances, such as, for example, a decline in the demand for men with his particular qualifications. Positive assortative mating may also come into play. Our empirical model specifies spouse labour-market participation equations within each household, endogenizing wages and allowing for random effects and correlations in spouses' unobservables. We conclude that the determinants of wife-sole-earner households are quite distinct from those for dual-earner households where she outearns him. The probability of observing the first seems to be more related to labour market difficulties of the husband, while the latter is not. Dual-earners where she outearns him are more likely to be found among higher educated couples, and especially, among couple where the wife's education level is high.