To systematically determine the effects of daytime exposure to sleep in darkness on human circadian phase, four groups of subjects participated in 4-day studies involving either no nap (control), a morning nap (0900-1500), an afternoon nap (1400-2000), or an evening nap (1900-0100) in darkness. Except during the scheduled sleep/dark periods, subjects remained awake under constant conditions, i.e., constant dim light exposure (36 lx), recumbence, and caloric intake. Blood samples were collected at 20-min intervals for 64 h to determine the onsets of nocturnal melatonin and thyrotropin secretion as markers of circadian phase before and after stimulus exposure. Sleep was polygraphically recorded. Exposure to sleep and darkness in the morning resulted in phase delays, whereas exposure in the evening resulted in phase advances relative to controls. Afternoon naps did not change circadian phase. These findings indicate that human circadian phase is dependent on the timing of darkness and/or sleep exposure and that strategies to treat circadian misalignment should consider not only the timing and intensity of light, but also the timing of darkness and/or sleep.