The obstetric and gynaecological histories of 80 women with proven placenta praevia have been reviewed together with case controls matched for age and parity. There was a significant relation between placenta praevia and previous caesarean section (P less than 0.05), dilatation and curettage (P less than 0.01), spontaneous abortion (P less than 0.05) and evacuation of retained products of conception (P less than 0.05). Repeated uterine instrumentation was associated with increased risk of placenta praevia (P less than 0.001). We were unable to show any influence of previous termination of pregnancy. These findings are consistent with the hypothesis that endometrial/myometrium damage is a significant aetiological factor in low placental implantation.