This case is reported not only because the transfusion of a group II (A) blood to a group III (B) recipient has been almost universally accompanied by severe reaction or death, but also because serologic study of the present case has shown why the usual fatal result did not take place. REPORT OF A CASE Mrs. M. H., aged 35, colored, weighing only 85 pounds (38.6 Kg.), was hospitalized for a pelvic operation. Her chief complaint was lower abdominal pain of three months' duration, especially on the right side. Because of a secondary anemia of 55 per cent hemoglobin with 3,400,000 red blood cells, a preoperative transfusion was deemed advisable. The patient's blood, with the use of the usual microscopic method of grouping, was found to belong to group III (B). Her husband's blood was grouped by the same method and appeared to be that of a universal donor; that