IT IS GENERALLY conceded that in the many cardiac clinics rheumatic heart disease is frequently overdiagnosed and underdiagnosed. The fact that this situation obtains is evidence of the inadequacy of the present diagnostic criteria used in most clinics. It is evident that the clinical manifestations of rheumatic fever have undergone a vast change of character. The magnitude of the shift is perhaps not generally appreciated and is not reflected in the textbook descriptions of rheumatic fever.1 These continue to describe classic attacks of rheumatic fever in terms of polyarthritis, fever, elevated sedimentation rate, arthralgias, and prolonged P-R intervals. In our experience and in the experience of most students of rheumatic fever, these cases are now rarely encountered. Major and minor criteria,2 when present, are of utmost diagnostic importance, but these are absent in the great majority of cases seen in the clinic. The "typical" case, in our experience,