Lithium has been tried in many psychiatric conditions but its efficacy has been firmly established only in the case of certain affective disorders. Many alcoholics are also clinically depressed and it therefore seems possible that lithium could be of benefit in treating alcoholics who have concurrent affective disturbances; this suggestion is supported by the results of studies so far reported.For the moment however, evidence from long-term investigations on alcoholics indicate that therapeutic effects of lithium are not simply caused by an alleviation of associated affective disorders, and the role of the mood disturbance in treatment response is not clear. Observations in acute studies suggest a possible mechanism of action, as lithium can attenuate the moodelevating effects of alcohol in subjects who abuse drugs or are depressed. Unfortunately many of these clinical reports are uncontrolled and can be criticized on other methodological grounds, and further double-blind studies which include comparisons with other drug treatments are required.Studies on rodents show that alcohol drinking is often reduced by lithium, but these results must be interpreted cautiously because lithium may produce conditioned aversion to alcohol and such a mechanism of action is unlikely in long-term studies on man.