In a multicenter 1-year trial of contraceptive vaginal rings (rings) involving 150 women, three dose combinations of the progestin Nestorone (NES) and ethinylestradiol (EE) were compared with respect to effectiveness, safety and acceptability. Mean in vitro drug release rates for the three doses were 150 and 15, 150 and 20 and 200 and 15 microg/day of NES and EE, respectively. Each ring remained in situ for 21 days, removed for 7 days and then reinserted for a total of 13 cycles of use. We studied ring performance with respect to pregnancy and other termination events, adverse events, the extent of ovulation inhibition, serum drug levels and bleeding control. We also assessed the rings' effects on the vagina using a standardized colposcopy procedure. Seventy-two percent of the women completed the 1-year (> or = 350 days) study. In studied cycles, luteal activity (progesterone > or = 10 nmol/L) was noted in 17%, 7% and 12% of subjects with monitored cycles at the 150/15, 150/20 and 200/15 doses, respectively (p = .34). Two pregnancies occurred, both in subjects using the 200/15 microg/day ring. Breakthrough bleeding during ring use averaged about 2 days/year and breakthrough bleeding and spotting averaged about 7 days/year. In the entire trial, only two women discontinued because of bleeding problems. Medical conditions, chiefly vaginal problems, personal reasons and device loss or repeated expulsion were the principal reasons given for study discontinuation. Vaginal and cervical colposcopy, conducted with standardized techniques and standardized interpretations, revealed no elevated event incidence attributable to ring use. Clinical performance and adverse event profiles indicate that each of these 1-year NES/EE rings, used on a 21-day-in and 7-day-out regimen, provided women effective, acceptable and safe long-acting contraception under their own control.