The effect of aging on the number of non-growing follicles (NGF) and early-growing follicles (EGF) was studied in humans through use of a database obtained by pooling two subsets of ovarian pairs (2 x 43 pairs) collected in two distinct populations. A previously suggested model of exponential regression of NGF counts in relation to the subject's age was tested but did not adequately fit the observed data points. This lack of fit is attributable mainly to the existence of a significant relation between a woman's age and the corresponding NGF count decay rate. Consequently, various regression models were tested. Two different periods of decay rate were observed for each population of small follicles. The first corresponds to younger ages with a decay rate that is slow for both types of follicles, although faster for NGF than for EGF. The second period corresponds to older ages with an accelerated decay rate that appears similar for NGF and EGF. The changing points were found at 38.0 +/- 2.4 and 39.0 +/- 1.9 yr (mean +/- SD) for NGF and EGF, respectively. Extrapolation of the fitted model suggested the presence of approximately 402,000 healthy NGF per ovary at birth and a total exhaustion of the follicular stock at around 74 yr of age. These results support the view that depletion of the NGF pool is caused mainly by atresia in younger women but mainly by entrance of NGF into the growing pool in older women. The mechanisms triggering accelerated entrance into the growth phase of NGF are discussed in relation to the previously reported increase in FSH plasma levels that starts in the late thirties, approximately, and precedes the menopausal period by several years.