A multicenter, international, randomized, comparative trial was conducted to assess the acceptability, efficacy and safety of two different schedules of a contraceptive pill, containing 250 micrograms levonorgestrel and 50 micrograms ethinyl estradiol, administered by the vaginal route. One schedule of daily administration for 21 days with a seven-day interruption to allow withdrawal bleeding was compared to daily administration without interruption for bleeding. A total of 900 women were recruited in three countries, Brazil, Egypt and China; 7,090 women-months of vaginal pill use were recorded (3,364 using the pills intermittently and 3,726 continuously). Four undesired pregnancies occurred, one in Egypt and three in China, all four in women using the pills intermittently. There was a statistically significant difference (p = 0.486) in pregnancy rate between the two groups. There were no other significant differences in discontinuation rates despite marked differences in bleeding patterns, amenorrhea predominating in the continuous use group. Hemoglobin levels increased significantly in the two groups but hematocrit was significantly higher in the continuous use group.