The purpose of this study was to compare tamoxifen to methotrexate, with respect to effectiveness, when followed by misoprostol to induce abortion. In the first phase, 198 women presenting for medical abortion at <7 weeks gestation were randomized to receive either 40 mg of tamoxifen, followed 2 to 3 days later by 800 mu g of misoprostol self-administered vaginally or 50 mg/m(2) of methotrexate, followed 5 to 7 days later by the same dose of misoprostol. In the second phase, 200 women were randomized to receive 20 mg tamoxifen twice daily for 4 days, followed by 800 mu g misoprostol or the same regimen of methotrexate and misoprostol as in phase I. The main outcome measure was success rate determined by the number of women who aborted without surgery. In phase 1, the success rate was higher in the methotrexate group (93.0%) compared to the tamoxifen group (85.7%) (p = 0.045). In the tamoxifen group, nine of 98 women had incomplete abortions with symptoms requiring a surgical aspiration, compared to one of 100 women in the methotrexate group. In phase 2, the success rates were 90.9% in the methotrexate group compared to 84.7% in the tamoxifen group (p = 0.20). The side effects were less in the tamoxifen group in phase 1 but not in phase 2. When tamoxifen is given as a single dose, it is less effective than methotrexate but when it is given 20 mg twice daily for 4 days, there is no significant difference. Tamoxifen does not appear to have any benefits over methotrexate. CONTRACEPTION 1999;59:265-270 (C) 1999 Elsevier Science Inc. All rights reserved.