A group of 120 women with gestations from 64 to 84 days received 800 micrograms of vaginal misoprostol every 24 h for a maximum of three doses without performing postexpulsion systematic preventive curettage. Outcome measures included successful abortion (complete abortion without requiring a surgical procedure), side effects, and mean time of expulsion and vaginal bleeding. Complete abortion occurred in 104 of 120 (87%, 95% CI 79, 92) subjects. The decrease of hemoglobin was statistically significant (p = 0.0001) but clinically unimportant: 12.2 mg/dL (SD 1.1) before treatment and 11.6 mg/dL (SD 1.0) after treatment. Statistically significant differences were found only between the success rates for white women in comparison with nonwhite women, in which case the success rates were higher for white than for nonwhite women. Vaginal bleeding lasted 8 +/- 5 days, spotting 4 +/- 3, and total bleeding 12 +/- 4 days. The acceptable expulsion time, the fact that postabortion systematic curettage was not needed, the clinically insignificant hemoglobin loss, and the success rate obtained show that misoprostol administered vaginally may be a valid method for interrupting gestations of 10-12 weeks.