MOST physicians who have used group psychotherapy are enthusiastic about it as a therapeutic tool. Although much has been written on the subject of group therapy in the past thirty years,1 very few writers have attempted to analyze their results statistically. Schilder2 and Bierer3 are exceptions. There are good reasons for this omission, and they apply to all attempts to analyze results from psychotherapy. Psychotherapy, in spite of the fact that it is on a more scientific basis now than ever before, still remains a highly subtle process, and each therapeutist develops his own individual skills. These do not lend themselves to ready analysis. Furthermore, the evaluation of therapeutic results is also a highly individual process. Results tabulated by one therapeutist are always open to serious doubt by others who do not know the exact criteria used to evaluate the results. What one person may call a