Oral administration of the progesterone receptor antagonist mifepristone, as well as the synthetic prostaglandin analog, misoprostol, have been shown to reduce cervical resistance to dilation in connection with legal abortion. The present study is a prospective, randomized study of 45 women comparing the clinical efficacy and possible modes of action of these compounds to induce cervical ripening by studying leukocyte populations and collagenolysis in the cervical stroma. A significantly increased amount of cervical dilatation, estimated to be the diameter of the largest Hegar dilator that passed the internal os without resistance, was observed after both mifepristone (200 mg, 48 and 24 h before surgery) and misoprostol (600 mu g, 16-20 h before surgery) compared to placebo, with no significant difference between the two preparations noted (mean 6.9 and 5.9 mm vs 4.5 mm, p <0.01). However, in the mifepristone-treated group of women, there was a higher proportion of cervices graded as easily dilatable compared with placebo- and misoprostol-treated women (26% vs 0% and 13%, p <0.05). As assessed by visual analog scale, the pain experienced during preoperative treatment by both substances was comparable and low (1.8 and 2.7). Bleeding starting before evacuation was noted in both of the treatment groups, but not in the control group, with a higher frequency in the misoprostol- than in the mifepristone-treated group of women (67% vs 27%, p <0.05). Side effects, such as diarrhea, were significantly increased after misoprostol as compared with mifepristone and placebo (20% vs 0% and 0%, p <0.05), but the frequency of vomiting did not differ significantly between the groups (20%, 13%, and 7%, respectively). Active pretreatment did not influence cervical gross morphology or stromal leukocyte infiltration as studied by immunohistochemistry. Neither was tissue collagenolytic activity nor plasminogen activator activity levels affected by mifepristone or misoprostol pretreatment. In conclusion, the study demonstrates a significant amount of increased cervical dilatation with mifepristone and misprostol, and indicates that this, as assessed by the present techniques, is not due to invasion of leukocytes or increased activity of collagenase or plasminogen actrivator. (C) 1998 Elsevier Science Inc. All rights reserved.