The objective of this study was to test the hypothesis that older reproductive aged women ovulate at a smaller follicle diameter and are more likely to produce multiple follicles during their menstrual cycle compared with midreproductive aged women. We performed a comparative study of 16 midreproductive aged women (MRA; 22-34 yr old) and 34 older reproductive aged women (ORA; > or = 45 yr old). Women underwent serial transvaginal ultrasounds to follow follicular growth over 1 menstrual cycle. A subset of women (nine MRA and 19 ORA) had daily blood sampling. Scans were initiated within 1 wk of menses and were performed at least 3 times/wk until evidence of follicular collapse was observed. If there was no evidence of follicle growth beyond 10 mm by 20 d, observations (ultrasounds and blood sampling) were ended. Follicle growth was organized backward from maximum presumed preovulatory diameter. Hormones were standardized to d 0, the day when progesterone levels exceeded 2 ng/ml. Group comparisons were performed using ANOVA with Mann-Whitney post hoc testing and Kruskal-Wallis testing for integrated hormones. The main outcome measures were peak follicle diameter, follicle growth patterns, and circulating LH, FSH estradiol, progesterone, inhibin A, and inhibin B. Six of 34 ORA women never underwent serial ultrasound. An additional 5 ORA women failed to ovulate on the basis of daily blood sampling or had no evidence of follicle growth beyond 10 mm by 20 d. Two of 16 MRA women were excluded: 1 due to severely decreased ovarian reserve at screening and 1 due to failure of follicle growth by cycle d 20. Small follicle counts in the follicular phase of the cycle (beginning of cycle through d -4) were greater in MRA women compared with ORA women (4.7 +/- 0.56 vs. 3.4 +/- 0.34; P = 0.042). Among presumed ovulatory cycles, ORA women demonstrated considerably more variable follicle growth patterns, with larger initial follicle size, but a trend toward smaller peak follicle diameters (15.22 +/- 0.95 vs. 17.85 +/- 0.71 mm; P = 0.07). ORA women were twice as likely to have multiple follicles as younger women (odds ratio, 2.06; 95% confidence interval, 0.93-4.6), but this observed difference was not statistically significant (P = 0.083). Comparisons of LH, FSH, estradiol, and progesterone between ORA (n = 14) and MRA (n = 8) women indicated the expected increase in FSH secretion, most evident in the early follicular phase of the cycle. Estradiol and progesterone concentrations did not differ between these groups. Inhibin B was decreased in ORA women compared with MRA women (P = 0.030). Despite normal-appearing patterns of follicle growth, grossly abnormal hormonal patterns were observed in some of the ORA women's cycles. Other cycles demonstrated a failure of folliculogenesis. These patterns are not observed in MRA women's cycles. ORA women ovulated at a smaller mean follicle diameter and had larger initial follicle diameters than younger women. The overall follicle growth curves of the older women tended to be flatter than those of the younger women. Taken together, the data suggest that follicle growth begins earlier in the cycle of perimenopausal women, but growth progresses more slowly. Ovulation may occur at an earlier stage of growth in association with reproductive aging.