A comparative multicenter clinical trial of two combined oral contraceptives (OCs) differing only in the estrogen content (35 mcg ethinyl estradiol versus 50 mcg mestranol) was conducted at five clinics located in Yugoslavia, Egypt, Sri Lanka, Costa Rica and Mexico. The trial was designed to determine the differences between Norinyl 1 + 35 (Syntex) and Norinyl 1 + 50 (Syntex) in rates and reasons of discontinuation, and frequency of selected side effects which might contribute to method discontinuation. This report includes analysis of 1698 women, all of whom were interval patients (at least 42 days but within 26 weeks postpartum), randomly allocated to one of the above OCs between October 1982 and January 1984. Follow-up visits were scheduled at 1, 4, 8 and 12 months after admission. Significantly more women in the Norinyl 1 + 35 group (p < .001) reported intermenstrual bleeding (primarily staining and spotting), as well as an increase in the occurrence of intermenstrual bleeding compared to women in the Norinyl 1 + 50 group. There were no significant differences between the groups for side effects with the exception of more women in the Norinyl 1 + 50 group (p < .05) reporting breast discomfort. The lost to follow-up rate at 12 months was 19.3% for both the Norinyl 1 + 35 and the Norinyl 1 + 50 groups. The total discontinuation rate (including women lost to follow-up) at 12 months was 43.5% for the Norinyl 1 + 35 group and 41.0% for the Norinyl 1 + 50 group. There were no significant differences between the two groups for gross cumulative life table discontinuation rates (p > .05). There were six accidental pregnancies attributed to user failure reported during the study period; four in the Norinyl 1 + 35 group and two in the Norinyl 1 + 50 group.