The aim of this study was to determine the incidence, indications, associations and complications of obstetric hysterectomy in our centre with a view to suggesting ways of improving outcome. This is a descriptive study of patients who had obstetric hysterectomy in the University of Benin Teaching Hospital from January 1986 to December 2000. Data obtained from hospital records were analysed using appropriate computer statistical software for graphs and chi(2) test. In the period under study there were 20 344 deliveries. There were 46 obstetric hysterectomies, resulting in an incidence of 0.226%. Incidence increased with increasing maternal age from 0.10% at 20 years to 0.706% at 40 years and above. Incidence also increased with parity from 0.028% at para 0 to 1.298% for grandmultiparous women. The indications for the procedure were ruptured uterus 14/40 (35.0%), atonic postpartum haemorrhage 13/40 (32.5%) and placenta praevia 9/40 (22.5%); 12/40 (30.0%) of patients had previous caesarean sections. Subtotal hysterectomy had been undertaken in 43.5% of cases. There was a maternal mortality rate of 12.5%. Mortality was associated with torrential continuing haemorrhage and septicaemia. Urinary tract injury occurred in 7.50% of patients. Injury to the urinary tract was associated with previous caesarean section and total hysterectomy. Obstetric hysterectomy can have grave consequences. Availability of blood products, intensive monitors and limitation of family size is advised.