Circadian rhythms have not been widely studied in schizophrenic patients, probably for methodological feasibility reasons. The available findings relate to the awake/asleep cycle and are not particularly up to date. Clinically' the reduction in overall sleep time and sleep fragmentation are usual but relatively non-specific findings in decompensated psychotic episodes. Although H Ey's hypothesis of intermediate sleep by has not been followed up, EEG recordings support moderate abnormalities linked to certain times of progression such as a possible increase in MOR density or reduction in REM latency in delusional phases. The reduction in SLP is thought to be more a marker of the schizophrenic state as an illness. A possible action of some antipsychotics on sleep fragmentation would give them resynchronising activity, which remains to be confirmed.