During the 1970s, information obtained from animal tumor models and from patients with a spectrum of solid tumors indicated the worth of a variety of immunostimulating agents. These findings provided a biological and clinical rationale for conducting randomized trials to evaluate the worth of those agents. Consequently, in May 1977 the National Surgical Adjuvant Breast and Bowel Project (NSABP) implemented a randomized trial to determine whether Corynebacteriurn parvum (C. parvum, CP) plus chemotherapy would be more effective than chemotherapy alone in prolonging the disease-free survival (DFS) and survival (S) of patients with primary operable breast cancer and positive axillary nodes. The results of that trial through 8 years of follow-up fail to indicate that treatment with CP used in conjunction with 1-phenylalanine mustard (L-PAM) plus 5-fluorouracil (PF) results in a better DFS and S than that observed after chemotherapy alone. Use of the immunomodulator has instead resulted in a poorer, but not statistically significant, outcome. Despite adjustments made to account for any imbalance in distribution of prognostic factors between the two treatment groups and despite considering treatment compliance as a factor, the unfavorable outcome persisted. A high incidence of fever and chills was associated with the administration of CP. The administration of hydrocortisone before each CP treatment reduced the frequency of those and other systemic effects. The failure to demonstrate a benefit from CP is in keeping with the failure of other nonspecific stimulating agents to contribute to the creation of a new paradigm for the treatment of breast cancer. Cancer 66:220-227.1990.