This is a prelimianry report of an effort to treat women with advanced (Stage III and IV) ovarian cancer who had progressive disease in spite of previous surgery, chemotherapy and/or radiation by a program of reductive surgery, intensive immune stimulation and combination chemotherapy. An initial laparotomy was done where possible to reduce tumor burden, and then all patients were given intravenous corynebacterium parvum (C.P.) in escalating doses over a 10- to 14-day period. Cyclic chemotherapy with Cytoxan, adriamycin and 5-fluorouracil (CAF) was started and repeated monthly. Maintenance subcutaneous C.P. was given weekly. All patients had frequent follow-up clinical and laboratory examination. Immune function was measured by skin tests and in vitro tests prior to treatment and periodically during therapy. Two-thirds of the patients had depressed DNCB and PHA stimulation responses prior to treatment, and almost all had severely depressed lymphocyte counts. Thirty-nine patients entered the program. Exploratory laparotomy was done in 16 patients and in eight, successful tumor reduction was accomplished. Eleven patients received intravenous C. Parvum and all expired before receiving chemotherapy. Four patients received C. Parvum and less than three cycles of CAF; all expired within 2 months. Twenty-four patients received C. Parvum and less than or equal to three cycles of CAF. Four patients had complete regression of measurable disease and were living free of disease 9-12 months after starting CAF. Eight patients had more than 50% regression for a minimum of 3 months, and five were living with disease (LWD) 5-11 months. Five patients had 25% to 50% regression and three were LWD 4-8 months. Seven patients had no regression and all expired within 4 months. Of eight patients who had successful reductive surgery prior to treatment, three were free of disease, median of 10 months, and five had partial responses and were living with disease, a median of 9 months. Although pre-treatment immune function was better in the patients who had a good response to CP and CAF (10 of 12 were DNCB+) vs that in patients with a poor response (4 of 12 were DNCB+) immune function was not significantly improved during therapy. The initial treatment results in this program are encouraging and suggest that this approach may be useful in patients with earlier disease.