To examine the relationship between the occurrence of menopausal hot flashes and the pulsatile release of LH, we have investigated the serum hormone levels and the occurrence of hot flashes by objective recordings in five women with endometriosis given daily injections of a long-acting GnRH agonist (GnRH-a) for 28 days. Results were compared to the findings made in 25 young women 6-8 weeks after bilateral oophorectomy. Serum levels of estrone and estradiol were similar in the subjects given GnRH-a and the women who underwent a surgical castration. In comparison with values before GnRH-a administration, the mean FSH level was lower whereas the mean LH concentration was significantly higher (P less than 0.01) on the last day of therapy. The coefficients of variations of both gonadotropins measured during 4-h sampling periods at 20-min intervals before and at the end of GnRH-a administration were significantly reduced (P less than 0.01) with therapy. During the total of 20 h of frequent sampling in the 5 subjects, 15 pulses (20% rise from nadir) of LH and 12 pulses of FSH were detected before GnRH-a, whereas only 2 and 8 pulses, respectively, were observed on day 28 of treatment. Hot flashes were observed in both groups of patients. The proportion of women experiencing hot flashes, the rate of occurrence/h and the characteristics of the physiological changes were similar in the 2 groups of women. These data indicate that hot flashes can occur in the absence of prominent LH pulses, suggesting the pulsatile release of this hormone is merely associated with the hot flash rather than being etiological.