A prospective randomized, double-blind, controlled clinical trial to compare the clinical efficacy and side effects of intravaginal misoprostol with the traditional prostaglandin, gemeprost, in second-trimester pregnancy interruption was conducted. A sample size of 100 women was calculated to demonstrate that misoprostol was as effective as gemeprost in achieving delivery within 24 hours (alpha = 0.1, 80% power). Women were recruited with fetal death in utero, severe fetal anomaly, or psychosocial pregnancy termination between 14 and 28 weeks gestation and randomized to receive either 1 mg gemeprost 3 hourly for 5 doses, or 200 mcg misoprostol 6 hourly for 4 doses, intravaginally. The therapeutic regimens were repeated if undelivered by 24 hours. Those undelivered after 48 hours received an extra-amniotic PGF2 alpha infusion. The median gestation at recruitment was identical: gemeprost 19 weeks (IQ 17-22 weeks) vs. misoprostol 19 weeks (IQ 17-21 weeks), P = 0.887. Delivery within 24 hours occurred in 75.1% of women receiving gemeprost and 74.9% receiving misoprostol (P = 1.0). The median time from prostaglandin commencement to delivery was similar: gemeprost 13.7 hours (IQ 9.0-23.5 hours) vs. misoprostol 16.9 hours (IQ 10.3-23.5 hours), P = 0.769. A significant reduction in the incidence of vomiting in women randomized to misoprostol occurred (34% vs. 13.2%, P = 0.017). There was no significant difference in the incidence of maternal fever > 37.5 degrees C, nausea, diarrhea, or placental retention. A 200-fold pharmaceutical cost advantage was observed with the use of misoprostol compared with gemeprost. Intravaginal misoprostol performs as effectively as gemeprost in achieving delivery in the second trimester without increase in adverse effects and displaying a significant cost advantage.