The author was able to trace and re-evaluate 254 of 299 borderline patients who had been treated at a long-term psychotherapy unit at the New York Psychiatric Institute ten or more years earlier. He found that the female patients who in retrospect had met the DSM-III criteria for diagnosis of borderline personality disorder seemed, generally, to have had the best recovery rate. Patients who were functioning at a borderline level but who manifested eating disorders, substance abuse, severe phobias, or a concurrent affective disorder were relatively responsive to a psychoanalytically oriented psychotherapy but had significantly poorer outcomes than those with pure borderline pathology. For example, the former group had a 12.6% suicide rate, as compared with a 3.0% rate in the latter. High function at follow-up did not, however, always correlate with efficacy of an exploratory psychotherapy; some patients recovered and continued to do well after minimal treatment or as a result of other treatments. Exploratory psychotherapy seemed to benefit only a third of this study population.