Arterial plasma noradrenaline and adrenaline concentrations before, during and after an attack of pain, induced first by constant supine exercise and then by multistep atrial pacing, were determined in four patients with coronary occlusion disease and stable angina pectoris. An identical protocol was applied to a patient with atypical precordial pain (anxiety state) and normal coronary arteriograms. When compared, the results led to the following conclusions: 1) during supine exercise arterial plasma catecholamine concentrations, particularly noradrenaline, progressively increase, reaching highest values in temporal coincidence with the onset or the peak of pain, 2) during multistep atrial pacing-induced angina no significant changes of arterial plasma catecholamine concentrations are seen. These data, obtained from the same patients, further emphasize that the application of atrial pacing to the study of pathophysiology of angina pectoris and for evaluating antianginal drugs, especially if interfering with adrenosympathetic system activity, must be considered with caution.