The results of various in vitro analyses indicate there is an active immune response against antigens associated with human malignancies. This immune response apparently can be augmented by nonspecific immunologic stimulates such as BCG. These agents are effective for destroying tumor when injected locally into intracutaneous disease but are not as effective for subcutaneous disease. Preliminary clinical trials indicated that immune stimulants are effective when administered systemically. The effect is only minimal for diseminated disease, but the therapeutic benefit is clearly augmented for patients with a minimal residual tumor burden, such as those patients with metastases to regional lymph nodes. Thus immunotherapy is a systemically active mode of therapy. Its toxicity is minimal, and it appears to be effective in a wide spectrum of the disease. However, immunotherapy is not effective for a large residual tumor burden; consequently it must be used in combination with other modes of treatment such as irradiation therapy or chemotherapy. Early experiences with BCG immunotherapy for malignant melanoma and C. parvum for oat cell carcinoma are encouraging. It is remarkable that a nonspecific immunologic stimulant does, in fact, have this effect. Immunotherapy experiments in animals suggest that in order to achieve maximal benefit. BCG must have close contact with tumor cells or must be combined with a tumor-associated antigen. If these principles are true for man, it would seem that improvements for nonspecific immunotherapy in human neoplasms would be further augmented if a tumor-related antigen could be extracted from human tumours and combined with a nonspecific immunologic stimulant.