Risk factors for placenta praevia have been analysed in a case control study conducted in Milan, Italy. A total of 140 cases of placenta praevia and the two women who delivered healthy babies consecutively after each index case (280 controls) were identified reviewing clinical records of 49,765 registered deliveries at the Clinica Mangiagalli in Milan from 1979 to 1991. An increased relative risk (RR) of per cent confidence interval, CI, 2.3-6.7) and among those with previous abortions (RR = 1.8, 95 per cent CI, 1.2-2.8) or with more than one previous pregnancy (RR = 1.9, 95 per cent CI, 1.1-3.3). A non-significant trend in risk was observed with parity (RR = 1.2 for one and 1.5 for two previous deliveries versus nulliparas) and for caesarean section (RR = 1.2 for one and 2.1 for two previous caesarean deliveries versus no previous caesarean section). No relation emerged with sex of the newborn and multiple pregnancy and risk of placenta praevia. Our study shows that older age and previous abortions are associated with an increased frequency of placenta praevia.