A review of the literature dealing with the etiology of rheumatic fever shows clearly that much of the controversy in this important field is due to the paucity of adequate controls. It is a curious but undeniable fact that though bacteriologic investigations on this disease date back more than forty-five years, few of the numerous reports published are critical or well controlled. In order to establish a suitable background for the work to be reported in this paper, it may be well to mention briefly a number of the methods by which this problem has been attacked and to point out the doubtful justification in many instances for the conclusions drawn from the data obtained: The description of organisms found post mortem and their acceptance as the etiologic agent of rheumatic fever. This error dates back to the work of Birch-Hirschfeld,1and was repeated by numerous investigators. In a