The effect on labour of 50 mg intravaginal PG F2 alpha or a standard intravenous oxytocin regimen was compared in 2 randomised trials involving a total of 83 patients, 23 of whom had experienced spontaneous rupture of the membranes (S.R.O.M.) and 60 of whom had artificial rupture of the membranes (A.R.M.) to induce labour. In each trial, labour had not been initiated by membrane rupture alone. In both trials only 20% of the patients receiving PG F2 alpha required further augmentation of labour with intravenous oxytocin. The mean length of labour in patients receiving PG F2 alpha was 2.5 hours shorter in the A.R.M. trial and 3.0 hours shorter in the S.R.O.M. trial than the mean length of labour in patients receiving intravenous oxytocin (P < 0.01). In the A.R.M. trial, the PG F2 alpha-treated group had significantly less analgesic requirements (P < 0.001). Although more normal deliveries occurred in the patients treated with PG F2 alpha than oxytocin in both trials, the numbers did not reach statistical significance. No side effects occurred in the PG F2 alpha-treated patients or their babies and this method was much preferred by patients and nursing staff alike.