A prospective randomised controlled trial was undertaken to compare the efficacy of two routes of administration, oral versus vaginal, of the prostaglandin E1 analogue misoprostol (Cytotec) to effect termination of pregnancy in the mid-trimester. Fifty-five women were recruited into the trial; 26 to receive all doses orally and 29 via the vaginal route. The dosing regimen was 400 microg as the initial dose followed by a second dose of 200 microg two hours later and then four-hourly 200 microg doses until delivery or 32 hours from commencement of treatment. If delivery had not been effected by the last dose of misoprostol, a Syntocinon infusion was started synchronously Misoprostol administered vaginally was significantly more effective than when administered orally as judged by induction-to-delivery interval and also the need or otherwise to augment therapy with a Syntocinon infusion. The average induction-to-delivery interval was 17.5 hours in the vaginal group compared to 33 hours in the oral group (p = 0.0003). The percentages of women who delivered at 24 and 48 hours were 93% and 100% in the vaginal administration group and 19% and 70% in the oral administration group (p < 0.05). No significant differences in complication rates or side effects were noted between the two groups