Night and rotating shift work schedules induce misalignment between the phase of the circadian pacemaker and the work/sleep schedule, leading to excessive fatigue at work and disrupted daytime sleep in night workers. Also, night work often begins after more extended wakefulness, and hence greater homeostatic sleep pressure than daytime work. This combination leads to sleepiness, errors, and injuries on the night shift, as well as health problems. While treating subjects with bright light and a fixed morning sleep schedule results in rapid circadian adaptation to night work, there is currently a critical deficiency in understanding how different sleep schedules and light treatments combine to produce physiological and psychological adaptation to night work. High intensity bright light may be too expensive for some environments, and workers may be unable or unwilling to maintain fixed sleep schedules. Therefore, we plan to test the hypotheses: 1) that 600 lux will be as effective in shifting the circadian phase of subjects into an appropriate sleep episode as 2500 lux; 2) subjects on an evening sleep schedule will perform better and be more alert than subjects on a morning sleep schedule; 3) that subjects whose melatonin secretion peak occurs during their sleep episode will perform better and be more alert than subjects whose melatonin secretion peak occurs during the work shift; 4) given a fixed sleep schedule, subjects whose melatonin secretion peak occurs during their sleep episode will sleep better than subjects whose melatonin secretion peak occurs during the work shift; 5) the melatonin phase of subjects who maintain a fixed sleep schedule in the laboratory will be shifted further than that of subjects who sleep ad lib at home. We propose an experiment to evaluate the effects of 3 different intensities of appropriately timed bright light during the night shift in combination with either AM, PM, or free sleep schedules on circadian phase, sleep alertness, and performance of subjects in a high-fidelity laboratory simulation of night work. We propose to measure circadian phase using salivary melatonin levels; alertness and performance via computerized assessments; and sleep via polysomnography. This work has major implications for the health and safety of shift workers.