Summary An analysis of the clinical course exhibited by 225 rheumatic subjects who have lived to their present ages of sixteen to thirty years, and 112 individuals who have succumbed to the disease during a twenty-year period of observation, has been presented. The onset of the disease from childhood is characterized by protean manifestations of activity and by a varied number of years of freedom from recurrence or continuous years of manifestations of active infection. After puberty, about the age of sixteen years, the majority of the subjects (66 per cent) were free from recurrence of symptoms, compared with 13 per cent before this age period. The practical importance of consideration of the age factor in this disease has been discussed. From a comparison of the course of the disease in the living and in the deceased series of subjects, it appears that the disease is self-limited, death being due in most instances to fatal active carditis. While the first attack of carditis may be severe and fatal, the majority of the deceased subjects survived subsequent recurrent active carditis. The average age of death was 12.6 years, and the greatest incidence was in the age group of twelve to sixteen years, inclusive, for the series. The care and management of these patients during the period of observation is not within the scope of this report. It is, however, of some interest to note that the children who experienced several intercurrent years of freedom from manifestations of rheumatic activity, remained in the same environment. Those who succumbed to the disease and those who survived with minimal degree of cardiac damage received comparable medical supervision. It is difficult to reconcile some current etiologic hypotheses with the observed course of the disease in these subjects. The type of rheumatic manifestation, exclusive of active carditis, did not appear to be of prognostic significance. The incidence of heart disease was comparable whether patients exhibited joint pains alone or chorea or polyarthritis with or without joint pains. All of the subjects presented evidence of cardiac involvement, although in only 39 per cent were the symptoms of carditis above the clinical horizon and capable of recognition. When the symptoms of active carditis were of sufficient degree to be diagnosed, the subsequent amount of cardiac damage, that is the multiple valvular lesions and marked cardiac enlargement, was greater and the mortality higher. During the course of the disease physical signs of valvular involvement may disappear, may be inconstant or uncharacteristic, leading to the erroneous opinion that the heart is left unscathed. The presence of considerable cardiac enlargement, established by adequate roentgenologic examination in a large series of these children who were under continuous careful observation, indicates the limitations of a diagnosis of the absence of cardiac involvement in rheumatic individuals based on physical examination alone. It is of some prognostic importance to note that the majority of these rheumatic individuals who survived were able to carry on their work and play without circulatory symptoms, particularly after puberty. The future course of the disease in these individuals must await a further period of continued observation for final comment. It would seem likely, however, that the usual unfavorable prognostic opinion that the majority succumb during the third and fourth decades is influenced by observations which are limited to adult patients who are under observation because of symptoms or by postmortem statistics. There may well be in the general population a considerable number of rheumatic individuals in whom the disease has been mild or unrecognized and the presence of heart disease unknown. Our own observations, as well as the frequent accidental discovery of mitral stenosis during routine physical examination or postmortem examination, favor this view. Accumulation of adequate data as to the true incidence, morbidity, and mortality of this disease is a vital need. Such information can only be collated from studies of representative series of subjects under continuous observation from the onset of the disease throughout its course to its end. An adequate comprehension of the natural history of rheumatic fever is the only reliable criterion available at the present time for considering critically current etiologic hypotheses and for evaluating various therapeutic measures.