This article reports the first of a series of studies designed to provide information about some of the relationships and consistencies among the Szondi test items. The Szondi Test is a relatively recent contribution to the area of projective techniques and psychodiagnostic tests. It consists of 48 pictures of individuals (of both sexes) representing eight different pathological categories. 2 For purposes of administration the 48 pictures are divided into six sets of eight pictures each. Each set includes a picture representing each of the eight diagnostic categories. Essentially, the task of the subject is to choose from each set of eight pictures the two pictures he "likes" and the two he "dislikes." Choices from all sets are considered when interpretations are made. However, this study was not concerned with the test as a diagnostic device, and it is not to be interpreted as a "validity" study. For this reason more elaborate details of the theory and use of the Szondi Test are omitted and the reader is referred to Deri [2] for such a treatment.Effective utilization of this device would seem to require that the test show a kind of internal consistency. That is, it would seem important that the various "items" used for the same purpose be homogeneous and consistent in their function. Thus, the design and nature of these studies are intended to provide information of a fundamental kind about the 1 Mr. Patterson is now at Baldwin-Wallace College. The collection and tabulation of the original data were carried out by Mr. Patterson and are included in an unpublished master's thesis at the University of Nebraska entitled, "A Study of Matching Behavioral Descriptions with Szondi Pictures." This work was carried out in 1949-50. The data have since been reanalyzed by Mr. Dudek and it is this reinterpretation that is presented here.