Reviewers of projective tests no longer find them "promising" (Eysenck, 1959; Shaffer, 1959). This criticism may reflect the absence of an acceptable frame of reference within which the research literature may be evaluated. Three postulates about validity, taken from one such frame work (Dana, 1960a), are considered here. A. Projective tests provide data from which there is a high probability of making accurate predictions about human beings. B. Projective tests provide behavior samples which are positively correlated with external observations of the same behavior. C. Projective tests elicit data that are positively related to theoretically derived constructs which are descriptive of human beings.