Introduction - Depression, by virtue of its conceptual and/or etiological frameworks, is an emblematic disease of new approaches that have been developed in social epidemiology: life-course epidemiology and contextual epidemiology. The objective of this study is to propose hypotheses for understanding this disease using these two approaches in a complementary way. Its goal is to look for an association between potentially traumatic life events during childhood and the occurrence of a depressive state during adulthood, and if, after adjustment for individual characteristics associated with depression, differences persist according to the type of neighborhood of residence. Methods - The SIRS study has been following a random sample of 3,000 individuals representative of the adult French-speaking population of Paris and its inner suburbs since 2005. The present analysis uses individual baseline data gathered through face-to-face interviews concerning, among other things, the subjects' depressive state, as evaluated by the MINI-Diag, life events and social conditions during childhood. The respondents' type of socioeconomic census block of residence was taken into account as well. The analyses use logistic and multilevel regression models. Results - After adjustment for the subjects' demographic and socioeconomic characteristics, several life-course events during childhood and certain characteristics of family functioning appeared to be strongly and independently associated with depression on the day of the survey, in particular, having been a victim of sexual abuse or having witnessed interparental violence, but also the parents having experienced long periods of unemployment or serious financial difficulties. In addition, after adjustment for these individual characteristics, depression was still more frequent in the neighborhoods considered to be in vulnerable urban areas. Conclusion - Several different life events and conditions during childhood appear to be detrimental to mental health during adulthood, and territorial inequalities persisted after adjustment for these factors.