This study was designed to evaluate follicular activity in women taking oral contraceptives with imposed imperfect compliance. After completing a 28-day cycle of either triphasic norgestimate/EE (NGM/EE) (Ortho Tri-Cyclen, Ortho-McNeil Pharmaceutical, Raritan, NJ) or monophasic levonorgestrel/EE (LNG/EE) (Alesse, Wyeth-Ayerst Laboratories, Philadelphia, PA), women were instructed to intentionally "miss" the first two active pills of the next pack. The first two tablets in the second treatment cycle were deliberately omitted, thereby extending the pill-free interval from 7 days to 9 days. Subjects were randomized to take NGM/EE (n = 40) or LNG/EE (n = 39) for two consecutive cycles. The mean maximum follicular diameter was significantly greater in women taking LNG/EE than in those taking NGM/EE (16.4 +/- 7.1 mm vs. 12.6 +/- 8.3 mm, p = 0.047). The LNG/EE group had significantly higher median serum estradiol concentrations compared to women taking NGM/EE on pill Days 10 [29.5 pg/mL (range: 10.0-540.0 pg/mL) vs. 2.5 pg/mL (range: 2.0-6.0 pg/mL), p < 0.001] and 14 [11.0 pg/mL (range: 2.0-416.0 pg/mL) vs. 2.0 pg/mL (range: 2.0-3.0 pg/mL), p = 0.001]. Two women in the NGM/EE group and three women in the LNG/EE group had at least one progesterone level > or =3 ng/mL; none of these women demonstrated a maximum follicular diameter >13 mm. Significantly greater follicular activity was observed after an extended pill-free interval in women taking LNG/EE compared to those taking triphasic NGM/EE. The clinical implications of these findings require further study.