The Southwest Oncology Group has conducted a series of randomized studies of adjuvant therapy in patients with primary breast cancer and positive axillary nodes. The first study, during which combined chemotherapy with cyclophosphamide, methotrexate, 5-fluorouracil, vincristine, and prednisone (CMFVP) administered for 1 year was compared with single agent therapy with melphalan (L-PAM) for 2 years, was activated in 1975 and closed in 1978. Of the 366 patients who were eligible, 191 received L-PAM and 175 were given CMFVP. The 2 groups were comparable with regard to known prognostic factors. At a median follow-up of 8 years, CMFVP continues to be superior to L-PAM in disease-free (P = .005) and overall survival (P = .01). Thirty-five percent of patients on CMFVP have died compared with 46% on L-PAM. The greatest survival benefit is apparent in premenopausal women and women with 4 or more positive axillary nodes. Acute toxicity was more frequent with CMFVP than with L-PAM, but it was acceptable and reversible with both regimens. Long-term toxicity was limited to myeloproliferative disease in 2 patients on L-PAM and 1 on CMFVP. The second-generation series of studies, activated in 1979, were designed according to nodal status, estrogen receptor (ER) status, and menopausal status. We randomized ER-negative, node-negative patients to 6 months of combination chemotherapy or to no treatment. Because of slow accrual, the study was closed and patients are now being entered onto a similar intergroup study with the Eastern Cooperative Oncology Group. We randomized ER-negative, node-positive patients to either 1 or 2 years of CMFVP.(ABSTRACT TRUNCATED AT 250 WORDS)