The blood group to which an individual belongs is a proved genetic characteristic and not subject to alteration by acquired disease. One of the questions which has received a great deal of attention in the field of cancer research is whether a constitutional factor enters into the causation of malignant growths. Accordingly, it seemed desirable to investigate the possible existence of a link between blood groups and the incidence of malignancy. In 1921 Alexander (1), in a study of 50 patients suffering from cancer, found an increase in the percentages of groups A and AB. Hoche and Moritsch (2) found a slight increase in groups A and B (20 per cent) among patients suffering from malignant disease. They do not mention, however, the type or locality of the tumors in the cases investigated. Johannsen (3) observed that among 263 tumor cases in Denmark there was an increase in groups A and AB in comparison with normal individuals. Matousek (4) investigated 200 female patients suffering from malignant tumors and made the observation that in carcinoma of the breast, group O predominated (S1.7 per cent). On the other hand, Thomsen (5), from a study of 1200 cancer cases, came to the conclusion that there is no definite deviation from the distribution of blood groups between cancer-bearing and normal individuals. Mario Einaudi (6), from a comparative study of blood groups from normal persons and those suffering from various diseases, including 50 tumor cases, found no marked difference. Hirschfeld and Hittmair (7), in an investigation of 150 cancer patients, observed that the distribution of blood groups showed no variation from normal. Buchanan and Higley (8) did not find any marked difference between the blood groups in relation to different diseases, including cancer.