Postnatal blues have been regarded as brief, benign and without clinical significance. However, several studies have proposed a link between blues and subsequent depression but have methodological problems. We report a prospective, controlled study of postpartum women with severe blues which uses systematically devised and validated instruments for that purpose which tests the hypothesis that severe blues increases the risk of depression in the six months following childbirth. 206 first-time mothers were recruited in late pregnancy. Blues status was defined using the Blues Questionnaire and those with severe blues and their controls who had no blues (matched for age, marital status and social class) were followed for 6 months with postal Edinburgh Postnatal Depression Scale. RDC diagnoses were made following SADS-L interview at the end of the protocol. Backwards stepwise Cox regression analysis found severe blues and past history of depression to be independent predictors each raising the risk by almost 3 times. Depression in those with severe blues onset sooner after delivery and lasted longer. The difference was largely accounted for by major depression. Severe postpartum blues are identified as an independent risk factor for subsequent postpartum depression. Screening and intervention programs could be devised.