THE AUTHORS wrote in a REVIEW article three years ago (38) that research with projective techniques presented more of a challenge than research with conventional psychometric methods because there was no clear-cut agreement as to the rationale for the whole process, because it was extremely difficult to find reliable criterion measures, and because there was no common metric. As was indicated, due to these difficulties the researcher with projective methods often failed to employ scientific methods, failed to use control groups, used too few cases, tended to overgeneralize his findings, described his scoring procedures too vaguely, used ill-defined criterion measures, and continued to rework concepts that research had shown to be neither important nor meaningful. It seems to the writers that most of the research reports in the period covered by this review are still limited by such difficulties.One of the leading Rorschachers, Klopfer, described the current situation when he said that there seemed to be a dearth of carefully designed longitudinal studies, and the evaluation of published studies seemed to be a thankless and almost impossible task because of the multiplicity of scoring systems which are not mutually translatable (22). He also charged that findings are so influenced by the researcher's choice of method of administration and scoring that any comparison of results seemed almost impossible. Two major critical reports are worth special consideration. Cronbach and Meehl (8) stressed the need to use widespread negative evidence so often found in studies with projective techniques and suggested that these results might contribute in building psychological constructs which would add to the validity of projective techniques as a whole. Lindzey (29) pointed out that one of the factors which contributed to the slight progress made toward an understanding of the Thematic Apperception Test (TAT) was an excess of casual empiricism and a scarcity of systematic investigations.