A study was made of variations of the Q-Tc interval as a result of exercise on 40 subjects, all submitted to coronary-angiography. The exercise tests were carried out on an ergometric bicycle and the Q-T interval, corrected for heart rate according to Bazett's formula, was measured on the resting ECG before the start of the test, and on the ECG recorded immediately following the end of the exercise. Of the 40 subjects, 29 had coronary heart disease and 11 were considered normal or showed non-significant lesions. In these 11, except for one with left bundle branch block, the Q-Tc showed a shortening on the ECG following exercise. Of the patients with coronary heart disease with or without the presence of contractility disturbances, 20 showed a prolongation of the Q-Tc following exercise, and eight showed no change. The Q-Tc variations were 0.05 sec. +/- 0.03. Analyzing the data as a 2 X 2 contingency table, the value of the alpha 2 corrected for continuity is 30.4, which is highly significant. Our study thus revealed that Q-Tc variations with exercise have a value for the diagnosis of coronary heart disease in exercise tests.