In order to determine whether the pituitary-adrenal cycle of normal man is synchronized by the dark-light cycle or by some other phenomenon associated with the sleepwake cycle, normal subjects underwent experimental alteration of their dark-light schedules in such a way that partial dissociation of the darklight schedules from the sleep-wake schedule was achieved. In addition, 3 blind subjects were studied. The normal subjects were allowed several days to become adapted to a given schedule prior to sampling of blood for plasma 17-hydroxycorticosteroid (17-OHCS) determinations. Hourly blood specimens were obtained through indwelling venous catheters in such a manner as to avoid disturbing sleep or exposing the subject to light during dark periods. Each subject slept 8 hr each day from 10 PM till 6 am, and the schedule of meals and activity was identical for all studies; only the dark-light schedule was altered. When the hours of darkness were scheduled in the middle of the waking day, from 10 am until 6 pm, normal subjects exhibited, in addition to the usual 17-OHCS peak at the time of awakening, a second peak in plasma 17-OHCS at the time of illumination. When darkness began at 10 pm but was prolonged until 10 am, 4 hr after awakening, the peak in plasma 17-OHCS occurred only when the lights were turned on, not when the subjects awakened. When normal subjects were kept in total darkness except for 1 hr of light from 6 till 7 pm, the peak usually seen at the time of awakening was delayed and diminished and a second major peak was observed during the brief period of illumination. Two of the 3 blind subjects were found to have plasma 17-OHCS rhythms that were not normally synchronized with sleep. One blind subject, studied on 2 occasions, appeared to have a shift in the timing of the 17-OHCS cycle, suggesting that he had a freerunning circadian rhythm. It is concluded that the dark-light cycle may be an important synchronizer of the pituitary-adrenal cycle in man.