This chapter has presented a somewhat complex view of the gonadotrope population, indicating that it consists of independent subsets. There may be regulatory cells that influence development and other ancillary processes needed for normal reproduction. For example, normal differentiation of PRL cells requires a functioning population of gonadotropes (Kendall et al., 1991). In addition, gonadotropes appear to be autoregulatory; subsets may produce inhibin or activin (in rats) and follistatin. Production of GnRH itself may serve as another regulatory tool. The gonadotrope population appears to be quite dynamic and convertible in the female rat. Cytological and cytochemical changes with the stage of the cycle are obvious. Increases in the numbers of immunoreactive gonadotropes parallel increases in GnRH target cells and culminate in peak expression of LH and FSH beta subunit mRNAs. The immunoreactive gonadotropes are greatly reduced after the surge activity, as though the cells had disappeared from the population. However, gonadotropes can still be detected by their content of gonadotropin mRNAs. This finding has led to the hypothesis that the gonadotropes recycle themselves. However, do they go through a resting phase? Is there a normal cycle of cell death and turnover? These are basic questions that must be answered in order to understand how the population is organized and renewed. Finally, we have returned to one of our original problems. Whereas it is clear that nonparallel release can be brought about by granules or cells with only one gonadotropin, the exact mechanisms that sort the gonadotropin molecules or turn off bihormonal expression are not known. A combination of autoregulatory events involving follistatin, activin, inhibin, and possibly steroids may play a role in modulating expression by a given subset. Delays in maturation may also prevent secretion of FSH and, hence, effect the delayed rise seen during late proestrus. The nonsecretory FSH cells seen in the studies by Lloyd and Childs (1988a) may be delayed maturers, requiring additional receptor types or changes in the calcium flux pattern to secrete their product. We also have a new question to address. What is the significance of the presence of GH in proestrous gonadotropes? Is GH a regulatory hormone, bound to receptors inside gonadotropes, or do subsets of somatotropes augment the population, producing a cocktail of GH and gonadotropins to aid ovulation? Either hypothesis is intriguing. Co-storage of GH and gonadotropins would be an efficient way of providing the hormones needed by the ovary. However, further work with in situ hybridization is needed to detect GH mRNA in such cells.(ABSTRACT TRUNCATED AT 400 WORDS)