The work of Hill, Long and Lupton (1) on the effects of muscular exercise in normal men, and more recently that of Furusawa (2) on the carbohydrate metabolism during exercise, made it desirable to undertake a similar investigation of the reactions to exercise of a diabetic individual with his deficient power of utilizing carbohydrates. Furusawa’s results show that in normal men on an ordinary diet the respiratory quotient of the excess metabolism due to exercise is unity, for periods of moderate exercise up to about 20 minutes. For similar exercise of longer duration the respiratory quotient of the excess metabolism is lower. When the subject had been living for some time previous to the experiment on a diet consisting chiefly of fat, the respiratory quotient of the excess metabolism was still unity for exercise of very short duration, but fell below unity for efforts of much shorter duration than 20 minutes. In the diabetic the carbohydrate reserves are less than in a healthy man, and his diet is restricted as regards carbohydrate and contains an excess of fat to make up the requisite number of calories. The diabetic, therefore, under such conditions, might be expected to approximate, in regard to exercise, to a healthy man living on a diet largely composed of fat. In the present investigation the effect of muscular exercise has been studied in three diabetic patients on lines similar, to those adopted by the above authors. We hoped that by exercising these patients with and without injections of insulin, some conception might be formed of the rôle of insulin in the metabolic processes involved in muscle during and after exercise. Recent work by Himwich, Loebel and Barr (3), and by Doisey, Briggs, Weber and Koechig (4), has shown that in the diabetic animal the formation of lactic acid from the glycogen of muscle on exercise is independent of the presence of insulin. Our results on diabetic patients confirm this view. We desired also to ascertain the likelihood of the development of ketosis and diabetic coma as the result of exercise in diabetics, together with the modifications which insulin might have on this process; further, there was the question of restriction of exercise in such patients, if it acted deleteriously. We attempted, moreover, to obtain information as to whether the dosage of insulin required depended on the amount of muscular work which the individual undertook, and also as to the need for regulation of exercise after a patient had been discharged from the inactive life in hospital.