We aimed to identify the risk factors and subsequent maternal and perinatal outcome associated with the procedure of peripartum hysterectomy. This was a retrospective case review carried out at the National Maternity Hospital, Dublin, of all patients who underwent a peripartum hysterectomy from January 1990 to December 1999. Seventeen cases were performed during the study period from a total of 64 563 deliveries (0·03%). Mean maternal age was 34 years (21-43). There were no nulliparous patients. Eight patients had previous vaginal deliveries. The other nine had a history of previous caesarean section, four of whom had two or more caesarean sections. The indications for the hysterectomy were placenta praevia (5), secondary postpartum haemorrhage (3), abruptio placentae (3), atonic postpartum haemorrhage (2), placenta accreta (2), traumatic postpartum haemorrhage (1) and broad ligament haematoma (1). All patients received blood transfusions, 74% receiving more than 10 units. There were no maternal deaths. There were two neonatal deaths secondary to placental abruption (perinatal mortality rate 118/1000). In conclusion, there was no maternal death in this study but the associated maternal morbidity and perinatal mortality was high. More than half the cases were associated with previous caesarean section. The caesarean section rate is increasing worldwide for many reasons, including recent emphasis on maternal request. We recommend that caesarean section be performed only for valid clinical indications, which should help to reduce the problems associated with peripartum hysterectomy.