It is confirmed that a fluorodinitrobenzene (FDNB)-treated frog sartorius muscle does not split phosphorylcreatine in the course of its contraction cycle, but does use adenosine triphosphate (ATP).2. Good stoicheiometric relations between the diminution of ATP and the formation of adenosine diphosphate (ADP), adenosine monophosphate (AMP) and phosphate are obtained, and in a 0.2 sec tetanus at 0 degrees C the net break-down of ATP amounts to 0.27, the total equivalent break-down to 0.34 mumoles/g.3. There is no difference in this quantity between muscles interrupted at the height of contraction and those that have also relaxed, and, in experiments specifically designed to determine relaxation metabolism separately, no such metabolism is found. Thus, all the ATP-break-down occurs in the contraction phase.