Life-events occurring at a discrete point in time were by no means the only kind of adversity faced by the women in Camberwell. A woman living for three years in two small damp rooms with her husband and two children would not have been picked up by our measure of events unless her situation had led to some kind of crises in the year - say that plans to move to another flat had fallen through or she had become pregnant. There is clearly no theoretical justification for dealing with such situations only when they give rise to a crisis. To our knowledge one other research team had tackled the issue: Coates and his colleagues (1969) asked about relevant problems in a survey in Toronto, using a list developed by Elinson, Padilla, and Perkins (1967). However, other life-event studies usually include several items reflecting ongoing difficulties. The New Haven study for example, asked about 'major financial problems' and 'increase in arguments with spouse' and these accounted for as many as a quarter of the 'events' occurring to patients (Paykel, 1974: 139). But living in an overcrowded flat is not of the same order of experience as unexpectedly finding that your husband has not long to live; and it does not follow that a woman who has lived with distress and frustration for years will be at risk of depression in the same way as one who has just experienced major loss or disappointment. We therefore set out to develop a quite separate measure of ongoing difficulties and, as we will show, found certain types to be of aetiological significance.