Longitudinal analyses, which track particular individuals through time, are few, and look mainly at the evolution of the division of household tasks between partners. They emphasize the growth of the time that women devote to housework when the couple and the family form, but they do not say whether this phenomenon is reversible. Does the opposite occur at the end of the relationship? What is the pattern for men? Longitudinal analysis of the data of the Swiss Household Panel (SHP) shows that the end of conjugal relationships (due to separation or death) leads to a reduction of the time that women devote to housework, whereas it has little effect on men’s investment. A discussion of the various factors that may explain these results motivates us to nuance the explanatory range of the “doing gender” theory which is widely evoked in studies on the division of household tasks within couples. This theory seems more appropriate to explain the behaviour of women than that of men. The household involvement of the latter seems, indeed, to depend less on the people with whom they interact than on cultural factors such as normative references with regard to household task allocation and investment which are specific to each generation.