The study of the creatine content of human voluntary and left and right ventricular cardiac muscle, carried out by Seecof, Linegar and Myers, has been extended to include determinations of potassium and phosphorus. It is evident from the studies of Fiske and Subbarow and others that in resting muscle creatine exists largely as phosphocreatine. It is likewise evident that this compound must be bound to some base, and this would appear to be largely, if not wholly, potassium. At about the time we began our original study on the creatine of heart muscle, Calhoun, Cullen, Clarke and Harrison pointed out that a diminished potassium content of heart muscle was invariably present in failing ventricles and suggested that this change may have been a contributing cause of such failure. Linegar and Myers noted that the creatine concentration also drops in heart failure and would appear to parallel very closely the observations of the Vanderbilt investigators on potassium. It was then suggested that this drop in potassium and creatine may represent 2 phases of the same process. Studies have now been carried out on 72 human autopsy cases in which potassium and phosphorus have been estimated in addition to the creatine. Since the cases were complicated by wide variations in diagnosis, it seemed advantageous to group them according to heart weights, excluding cases with nitrogen retention, which may cause elevated creatine values. The data studied were compiled from 44 hearts ranging in weight from 200 to 825 gm., comparable changes being found in both left and right ventricular muscle. It was noted that as the heart weight progressively increases there was a diminution in the concentration of the 3 constituents, creatine, potassium and phosphorus, the ratio of the average decrease being 3 :2 :1 respectively.