To compare the pharmacokinetics of vaginal and oral administration of the prostaglandin E1 analogue, misoprostol. Twenty women received 400-micrograms doses of misoprostol either orally or as tablets placed in the vagina. Serum levels of principal metabolite, misoprostol acid, were measured at 7.5, 15, 30, 45, 60, 90, 120, and 240 minutes. The first ten women were pregnant and undergoing first-trimester abortions, and the last ten were not pregnant and had additional blood sampling at 360 minutes. We compared the pharmacokinetics of misoprostol acid after oral and vaginal administration. All 20 subjects completed the study. The maximum mean (+/- standard deviation [SD]) of misoprostol acid differed significantly between the oral and vaginal groups (277 +/- 124 compared with 165 +/d- 86 pg/mL, respectively; P = .03, analysis of variance), as did the mean +/- SD time to peak levels (34 +/- 17 compared with 80 +/- 27 minutes, respectively; P < .001) and areas under the misoprostol concentration versus time curve (mean +/- SD) up to 4 hours (n = 20,273.3 +/- 110.0 compared with 503.3 +/- 296.7 pg.hour/mL, respectively; P = .033) and up to 6 hours (n = 10, 300.0 +/- 103.3 compared with 956.7 +/- 541.7 pg.hour/mL, respectively; P = .029). The extent of absorption was highly variable among subjects in each group. There are significant differences in the pharmacokinetics of misoprostol administered by vaginal and oral routes that may explain the difference observed in clinical efficacy. Assuming that the pharmacologic effect of misoprostol is related to its concentration in the plasma, our observation of the prolonged serum concentrations in the vaginal group suggests that vaginal administration could be dosed at longer intervals than oral.