Eleven nonathletes and eleven athletes were exercised on a standardised Harvard step test, and the average rate of change in QRS amplitude in lead III of the electocardiogram associated with heart rotation and the average change in rate of heart beat were observed. 1. After the Harvard step exercise, the average rate of change in QRS amplitude in lead III of both groups increased. This was due to the clockwise rotation of the heart and was associated with respiratory movement. The diaphragm was inferred to remain for a while in a relatively more inspiratory position. 2. After the Halved step exercise, a high correlation between the recovery of the average rate of change in QRS amplitude in lead III and the average change in rate of heart beat was observed in the athletic group. 3. In the nonathletic group there was no significant correlation between the average rate of QRS amplitude change and the average rate of change of heart beat. 4. Athletes were assumed to be trained to ventilate quickly at their maximum ability, using deep descending movements of the diaphragm and other respiratory musclature. Consequently, the average in rate of heart beat also recovered quickly. 5. Nonathletes were inferred not to have been trained to adjust quickly to ventilate so efficiently with their diaphragm movement and other respiratory musculature, and are characterised by their longer time to complete recovery.