To determine the effect of dietary energy restriction on gonadotropins, we assayed LH and FSH in samples drawn at 10- and 60-min intervals, respectively, over 24 h from seven young women (mean +/- SE gynecological age, 7.7 +/- 1.2 yr) on day 9, 10, or 11 of two menstrual cycles. Cortisol was measured in samples collected at 30-min intervals. During the 4 previous days and the day of sampling, dietary energy intake was set at either 45 or 10 Cal/kg lean body mass.day in random order. Beginning 2 days before treatment, blood was sampled daily at 0800 h and assayed for T3, insulin-like growth factor-I, and insulin. Estradiol was measured in samples collected daily and at 6-h intervals on the day of frequent sampling. By the day of frequent sampling, dietary restriction had reduced T3 20% (P < 0.01), insulin-like growth factor-I 58% (P < 0.001), and insulin 54% (P < 0.001). Twenty-four-hour transverse means for LH (P = 0.3), FSH (P = 0.2), estradiol (P = 0.3), and cortisol (P = 0.13) were unaffected, but LH pulse frequency was reduced 23% (P < 0.01), especially during waking hours, whereas LH pulse amplitude was increased 40% (P = 0.05), especially during sleep. These results support the hypothesis that LH pulsatility depends upon energy availability in women, as it does in other mammalian species.