We compared bright-light durations of 6, 3 and 0 hours (i.e. dim light) during simulated night shifts for phase shifting the circadian rectal temperature rhythm to align with a 12-hour shift of the sleep schedule. After 10 baseline days there were 8 consecutive night-work, day-sleep days, with 8-hour sleep (dark) periods. The bright light (about 5,000 lux, around the baseline temperature minimum) was used during all 8 night shifts, and dim light was < 500 lux. This was a field study in which subjects (n = 46) went outside after the night shifts and slept at home. Substantial circadian adaptation (i.e. a large cumulative temperature rhythm phase shift) was produced in many subjects in the bright light groups, but not in the dim light group. Six and 3 hours of bright light were each significantly better than dim light for phase shifting the temperature rhythm, but there was no significant difference between 6 and 3 hours. Thus, durations > 3 hours are probably not necessary in similar shift-work situations. Larger temperature rhythm phase shifts were associated with better subjective daytime sleep, less subjective fatigue and better overall mood.