Because early memories related by patients in diagnostic interviews have already proved to be a source of considerable insight, we felt it important to search for a method of systematic analysis of these recollections. Ordinarily the use of this kind of data relies heavily upon intuition and impressionistic skill. To suggest an alternative method of handling early memories we have applied a thematic-configurational method of analyzing these recollections. In this paper we shall describe our approach and the theory of it with a preliminary clinical test and a discussion of its application. Langs et al.1in their method of studying early memories focused on discrete units of analysis, e.g., whether an early memory was pleasant or unpleasant, the number of persons present, etc. and on manifest content "themes," i.e., the presence or absence of injury, punishment, etc. These units and themes were subsequently