Psychological testing procedures may be utilized to predict outcome in patients treated with drugs, thereby supplementing other prognostic indices and helping to specify the personal characteristics of patients responding well and poorly to treatment. The appropriate experimental design is that of the usual prediction study in which a set of predictors, in this case scores on tests administered before therapy, are correlated with a criterion, for example, clinical ratings of improvement. Research on prognosis may lead to the efficient selection of patients for a given kind of therapy and a rational basis for choosing among various possible therapies for a particular patient, and it may provide methods for selecting homogeneous groups of patients so that one kind of therapy can be evaluated against others or against a control.Psychological procedures may also be utilized to trace the effects induced by a drag, whether or not the changes are considered therapeutically desirable. The appropriate experimental design is test-retest, changes in functioning being inferred from changes in test scores. The objective here is to study change rather than improvement, to answer the question: "What changes?" rather than "Who changes?". Research employing this design may seek to establish psychophysiological relationships, for example, by studying the behavioral effects of two or more drugs which selectively modify different neurophysiological systems.The purpose of this paper is to suggest suitable testing procedures of the kind usually referred to as objective and projective tests of personality. Only passing reference will be made to experimental psychological procedures and to tests of sensory, perceptual, and motor functions. Clinical ratings by observers will be discussed where they are relevant to the problem of the criterion, and where they may be used to supplement the information derived from tests. Methods and results of studies of the somatic therapies (electroshock, insulin, lobotomy, etc.) and psychotherapy will be reviewed with an eye to their relevance for the study of drug therapy.