A few years ago Everson and I assembled all the examples of spontaneous regression in the world medical literature from 1900 to 1960 and added numerous cases from expriences of our friends. Our figure was 176. We excluded squamous cell carcinoma of the skin, leukemia, Hodgkin's disease, and a large number of cases that did not fulfill the prerequisites of confirmed diagnosis and no significant treatment. The four most common examples of regression were carcinoma of the kidney (31), neuroblastoma (29), malignant melanoma (19), and choriocarcinoma (19); these constituted more than half the group. We did not require that the regression be permanent because it appeared that the explanation of temporary regression would be just as important as the cause of permanent regression. There was no proven specific cause of the regression, but the following mechanisms had a possible relationship: immunologic action, elimination of carcinogens, trauma (altering the antigen-antibody relationship), hormones, irradiation, infection and/or fever, and drugs or chemicals. The most applicable of these is elimination of the carcinogen. Immunologic reactions seem to offer the best explanation, and the potential of humoral immunity is more impressive than that of cellular immunity.