Data are presented which confirm many observations of the early epidemiologists and bring to date certain aspects of the epidemiology of acute rheumatic fever, mainly through studies of death records. There is at present no general agreement as to the causal agent, mode of transmission or means of prevention. In the U.S.A. there is evidence of a striking decline in number of deaths and a tendency to agreement with the records for England in the years when peaks are noted. The author discusses aspects of the census methods for recording deaths from acute rheumatic fever and related conditions and concludes that these records are collectively trustworthy for broad conclusions on a statistical basis. The mortality graphs of age, season, sex, race and other factors appear to be too characteristic and orderly to fail to carry conviction as to their substantial accuracy. When rheumatic fever is compared with scarlet fever, chorea, erysipelas, septicemia and puerperal fever, it is seen that the trends as well as yearly oscillations are alike. Certain years prove good for all; other years prove poor. This correspondence appears in the U.S.A. as well as in the English records, where it may be traced back as far as 75 years. A review of the epidemiology of acute rheumatic fever indicates from many angles that it is one of the communicable infections caused by a parasite gaining access to the body from without, and producing a disease picture in which age, sex, seasonal, racial and other characteristics are fixed and very close in their resemblance to the well-known epidemic diseases. Taken with the recently confirmed evidence of multiple cases in rheumatic families, we may well agree with Hirsch that "rheumatic fever deserves an assured place among the acute infective diseases.".