This review paper deals with the issue of shift rotation, i.e. the num
ber of consecutive shifts of the same type and the order of change bet
ween shifts. To minimize the disturbances of the circadian system and
the accumulation of sleep deficits, rapidly and clockwise-rotating shi
ft systems would seem to be preferable. Regarding consecutive night sh
ifts, a week of night shifts seems to be the worst system with regard
to performance and accidents. There is a lack of reliable data on the
effects of permanent vs. rotating-shift systems on alertness, performa
nce and accidents. This is also true for the comparison of forward (de
laying) and backward (advancing) rotating-shift systems, although the
former would seem to be associated with fewer problems. For both syste
ms, controlled longitudinal studies are necessary.