IN RECENT years there has been a growing interest in the development of prognostic criteria in mental disorders. For the most part, early attempts in this area were concerned with the identification of prognostic factors of a clinically observable nature (3, 7). Predicting the outcome of psychotic disorders from psychological test performance is, by comparison, a fairly recent development.Windle (11), in a review of the literature on the prognostic use of psychological tests, has observed that a majority of studies on the prognostic value of various psychological tests report that good test performance is positively correlated with outcome. However, he also noted that a number of studies report an inverse relationship between test performance and outcome. The apparent contradictions were resolved when it was demonstrated that good performance on the part of patients with an acute (short duration) psychosis, and poor performance on the part of patients with a chronic (long duration) psychosis were both related to favorable outcome (12,15).A test of the prognostic value of psychological measures for recovery from psychosis must take account of those conditions which intervene between test observations and outcome. It is pertinent, for example, to consider the relationship between therapy and outcome. An evaluation of the contribution of therapy to outcome must be based on the comparison of spontaneous remission rates with remission rates following the various types of therapy. There is some agreement that special therapies produce remission rates no higher than good custodial care (9, IS). However, Zubin (13) has found that special therapies hasten improvement in some cases, although in the long run the remission rate may be the same as that following custodial care. Windle (11) concludes that, with respect to prognosis, individual factors are probably more important than type of therapy.