Selective elevations of plasma follicle-stimulating hormone (FSH) levels are characteristic of some physiological conditions, such as the early stages of human puberty, and in some disorders of testicular function, such as idiopathic oligospermia. We tested the hypotheses that a slow gonadotropin-releasing hormone (GnRH) pulse frequency favors a selective elevation of plasma FSH and that this is influenced by the circulating steroidal milieu. We administered exogenous GnRH at frequencies of once every 90 min (q 90 min) and once every 240 min (q 240 min) to castrated prepubertal male monkeys who had received either empty (sham) or testosterone (T)-filled Silastic capsules at the time of castration. At the end of each experimental frequency period, mean plasma levels of luteinizing hormone (LH) and FSH were measured. Plasma T levels were also measured. Animals with T implants had plasma levels of this hormone that were in the adult range (approximately equal to 8 ng/ml), whereas those with sham implants had plasma T levels in the prepubertal range (less than or equal to 4 ng/ml). In animals with sham implants, mean plasma FSH levels were markedly elevated at the slower GnRH pulse frequency (39.5 +/- 3.6 ng/ml following GnRH q 240 min compared with 23.7 +/- 2.8 ng/ml following GnRH q 90 min). This selective FSH elevation was not apparent in animals with T implants. Mean plasma LH levels were similar (approximately equal to 8 micrograms/ml) at the two GnRH pulse frequencies, in both T-treated and sham-implanted animals.(ABSTRACT TRUNCATED AT 250 WORDS)