Psychotherapeutic relationships between 18 physicians and 109 schizophrenic patients were compared and contrasted when insulin was combined with psychotherapy and when psychotherapy was used alone, in relation to the outcome of treatment. The patients are divided into an A group (those whose physicians achieved an improvement rate of 70% or higher) and a B group (those whose physicians achieved an improvement rate of less than 70%). Our findings indicate that:1. Insulin combined with psychotherapy is associated with a numerical increase in improvement only in the B patients (from 34% to 82%). A patients show no numerical gain (the rate remaining at 82%).2. The higher numerical rate of improvement in the B group when insulin is combined with psychotherapy does not carry with it an increase in quality of improvement. Only one B patient reached the highest grade of improvement and in this case insulin was not used. Nor are qualitative gains associated with the use of insulin in the A patients.3. The A physicians, with or without insulin, achieve the best qualitative results (12 of 13 patients at the highest grade of improvement).4. The empirical findings of our earlier study of 14 physicians and 100 schizophrenic patients are confirmed, in that favorable outcome was again found most likely when the physician's Personal Diagnostic Formulation shows some grasp of the personal meaning of the patient's behavior; when he aims at assisting the patient in modifications of personal adjustment patterns and a more constructive use of assets rather than focusing on psychopathology; when, in his day-by-day tactics he participates with the patient in an active personal way; and when a trusting, confidential relationship develops between physician and patient. These styles of transaction were again found to be more characteristic of the A than the B physicians.5. B physicians more frequently use the tactical pattern of active personal participation when insulin is combined with psychotherapy (54%) than when psychotherapy is used alone (9%). B patients who receive insulin are thus more frequently the recipients of one of the more effective psychotherapeutic patterns in contacts with their physicians than those treated by the same physicians without insulin, and this may account in considerable part for their greater numerical improvement.6. The tactical pattern of active, personal participation which is particularly highly associated with good numerical and qualitative levels of improvement (94% over-all improvement in this series, 85% in the earlier study), was not used in the majority of their cases by either the A or B physicians. It is suggested that this mode of therapeutic participation with schizophrenic patients merits more extensive trial.