Background. The high prevalence of depression in developing countries is not well understood. This study aimed to replicate the previous finding of a high prevalence of depression in Pakistan and assess in detail the associated social difficulties.Method. A two-phase survey of a general population sample in a Pakistani village was performed. The first-phase screen used the Personal Health Questionnaire (PHQ) and the self-rating questionnaire (SRQ). A one in two sample of high scorers and a one in three sample of the low scorers were interviewed using the Psychiatric Assessment Schedule (PAS) and Life Events and Difficulties Schedule (LEDS).Results. A total of 259 people were screened (96% response rate). The second stage yielded 55 cases, of whom 54 had depressive disorder, and 48 non-cases. The adjusted prevalence of depressive disorders was 44·4% (95% CI 35·3 to 53·6): 25·5% in males and 57·5% in females. Nearly all cases had lasted longer than 1 year. Comparison of the cases and non-cases indicated that cases were less well educated, had more children and experienced more marked, independent chronic difficulties. Multivariate analysis indicated that severe financial and housing difficulties, large number of children and low educational level were particularly closely associated with depression.Conclusion. This study confirms the high prevalence of depressive disorders in Pakistan and suggests that this may be higher than other developing countries because of the high proportion of the population who experience social adversity.