PIP: Further evidence that there is a relationship between prolactin rise and sleep is offered by this study that delayed sleep onset and inverted the normal sleep-waking cycle in 4 normal patients (3 women and 1 man). The inversion schedule lasted 3 weeks, and at its conclusion a final 12-hour laboratory study was carried out measuring the plasma levels of human prolactin and growth hormone by radioimmunoassay. In the first female subject, mean prolactin levels (ng/ml) were, for baseline and after shift, 7 and 8.3; for case 2 (female) the values were 7.3 and 5.9; for case 3 (male) the levels were 6.1 and 6.1; and for the last case, the levels were 4.7 and 4.3, respectively. Values for growth hormone were (ng/ml) 1.5 and 1.5 for Case 1 for baseline and after shift, respectively; 1.1 and 2.5 for Case 2, respectively; 1.8 and 1.1 for Case 3, respectively; and 7.3 and 5.2 for Case 4, respectively. Prolactin release shifted immediately and completely with shifts of sleep onset of 3, 6, and 12 hours. Thus, it is concluded that the nocturnal rise is dependent on the occurrence of sleep, not on the inherent rhythm related to time of day.