LITHIUM carbonate is now widely used for the treatment of mania,ls2 the prophylaxis of recurrent mania and depression2-4 and, to an extent, in the treatment of selected depressed patients. 5- 7 There is a clear trend towards the increased use of lithium for more patients and for longer periods of time.s Since depressed and manic patients have a significant sleep disturbance and. as it has been suggested that there may be significant associations between these sleep changes and both the clinical syndromes and the antidepressant drug effects,s-13 it seemed important to determine the effect of lithium on sleep (both acutely and chronically). Further, if lithium, which is now being used chronically for many patients, were to adversely affect sleep, then these patients may experience long-term sleep alterations of unknown implications. It is also of interest to determine what, if any, are the correlations between lithium’s therapeutic and prophylactic effects and its influence on the disturbed sleep pattern of these patients. Depressed patients (as a group) have a reduction in total sleep time, more awakenings during the night and less stage 4 sleep. 11, 14- 21 In addition, Mendels and Hawkins17,22 have reported that depressed patients take longer to fall asleep and wake significantly earlier in the morning and have increased amounts of stage 1 sleep. Reports on the amount of stage 1 REM sleep obtained by depressed patients are conflicting. Some studies report a reduction in REM sleep during depression, *I48 17* 20, 23-25 others report normal or even slightly increased REM sleep time with depression26p 27 with an earlier REM sleep onset1s,26 and an increase in phasic REM sleep elements such as eye movements. l s Later reports by MENDELS et aLa 2s* 3o likewise reflect a relatively high percentage of REM sleep as well as a pressure to achieve REM sleep in some (but not all) depressed patients. Two previous studies31* 32 of the effects of lithium carbonate on sleep offer apparently contradictory results, although the different findings may be a function of the dose administered and the subjects studied. BREBBIA et al. 31 administered 750 mg/24 hr lithium carbonate for 8 days to three control subjects and to three manic patients with plasma levels ranging between 0.34 and 0.46 mEq/l. The plasma levels are below those regarded as being in the therapeutic or prophylactic range. 2,8 They reported no uniform change in sleep pattern associated with the drug. KUPFER et al.,32 on the other hand, reported a significant increase in delta wave sleep (stages 3 and 4 combined) and a significant decrease in stage 1