ObjectiveTo determine whether misoprostol (a prostaglandin El analogue) 400 μg orally (group 1) or 800 μg vaginally (group 2) will cause complete uterine evacuation in women with early pregnancy failure.MethodsTwenty subjects were recruited for a prospective, non-blinded, randomized clinical trial. Early pregnancy failure was diagnosed by transvaginal ultrasound examination; only women with a closed cervical os and minimal vaginal bleeding were enrolled. Subjects returned 24 hours after misoprostol administration for a transvaginal ultrasound examination. If the gestational sac was still present, the misoprostol dose was repeated and the subject returned again 24 hours later. Subjects who failed to expel the pregnancy were offered a suction curettage.ResultsTwelve and eight women were randomized to groups 1 and 2, respectively. Complete uterine evacuation occurred in three of 12 [25%, 95% confidence interval (CI) 1%, 50%] and seven of eight (88%, 95% CI 65%, 100%,P= .010) subjects in groups 1 and 2, respectively. Vomiting occurred in 30% and 13%, respectively, and diarrhea in 50% and 38%, respectively.ConclusionVaginal misoprostol 800 μg is more effective than oral misoprostol 400 μg for uterine evacuation of early pregnancy failure and may be an effective alternative to dilation and curettage.