The official statistics on GDP and the labour market exclude unpaid domestic services. Yet there are good theoretical reasons why historians should study unpaid work. This paper reconstructs annual estimates of time use in Sweden from 1950 to 2012 among women and men. It finds substantial convergence between the genders in time use from the 1960s to the early 1980s. During the pe-riod of inquiry, the gender difference in total working time vanished. The double burden for women did not increase when they entered the labour market. The reduction in the time women spent on unpaid work is explained about equally by the shortening of the total amount of unpaid work and by increasing male participation inhousehold chores. In 1950-1963, the reduction was explained mainly by the decline in the making and mending of clothes at home and the spread of domestic appliances. In the 1963–1984 period, instead, it was due chiefly to men’s greater participation in household work. These mechanisms were largely historically contingent, suggesting that it is impossible to single out just one factor to explain why Sweden today has less gender inequality than other countries.