A prospective inquiry of a largely working-class sample of women with children considers the effect of employment on risk of developing clinical depression. The hypothesis was that there would be a direct protective effect arising from employment once quality of other support was taken into account. In fact full-time working mothers were at high risk. This appeared to be explained by either prior work strain or a severe event involving 'deviant' behaviour on the part of husband/boyfriend or child. Neither factor was relevant for part-time workers. The severe events appeared to be particularly depressogenic for full-time workers because they represented either failure in the motherhood role or a sense of entrapment in an unrewarding work/domestic situation. However, those in part-time work had a low rate of onset compared with non-workers, and the difference appears to be related to non-working women feeling less secure about their marriages.