In the literature the effectiveness of lithium prophylaxis in bipolar affective disorders is generally supported. In this paper we have examined whether besides the bipolar disorders the unipolar and schizo-affective disorders also benefit from lithium prophylaxis. For that purpose a group of 104 outpatients to whom lithium was administerted has been studied more closely with an individual retrospective control method and the life-table method. If effectiveness was expressed on the basis of the number of episodes and hospital admissions, than it appeared that notably the groups of bipolar and schizo-affective disorders did benefit and that the unipolar disorders did not. Indications were also found that young patients and patients who had not been treated for too long reacted more favourably to the prophylaxis. The number of episodes preceding the prophylaxis and the absence of unipolar depression were found to be predictors of effectiveness.