A postal survey of all women who gave birth in Victoria in one week in 1989 was used to assess the contribution of birth events, satisfaction with care and social differences to depression after birth. The questionnaire was mailed to women eight to nine months after birth. The survey response was 71.4 per cent (790/1107) with underrepresentation of young women, single women and women of non-English-speaking background. Assessment of depression was made using the Edinburgh Postnatal Depression Scale, with a score of 13 or more indicating probable clinical depression. The point prevalence of depression was 15.4 per cent (95 per cent confidence interval 12.8 to 18.0). In the final logistic model two factors were associated with lower odds of depression: nonmetropolitan residence and maternal age over 34. Factors associated with increased odds of depression in the model were having a first child over the age of 34, assisted delivery (caesarean, forceps, vacuum extraction), bottle feeding, dissatisfaction with antenatal care, having unwanted people present at the birth and lacking confidence to look after the baby at the time of leaving hospital. When those factors were taken into account, being unmarried, being born overseas of non-English-speaking background, being dissatisfied with care in labour and after birth, and not having an active say in decision making, all of which were significantly associated with depression in univariate analyses, did not make any additional contribution to the model.