This study examined the exercise capacity of trained and untrained limbs in men with angina pectoris before and after 8 weeks of arm (n = 4) or leg (n = 7) physical training or a control (n = 4) period. Time to angina (mean ± standard deviation) increased 3.6 ± 2.7 minutes (p < 0.01) during trained limb and 1.6 ± 1.2 minutes (p < 0.01) during untrained limb exercise. Myocardial oxygen demand at angina estimated by the product of heart rate and systolic blood pressure did not change with training. At a constant subanginal work load, rate-pressure product × 10 −2 was reduced by 35 ± 22 (p < 0.001) during trained limb and by 18 ± 27 (p < 0.05) during untrained limb exercise. The decrease in rate-pressure product with both trained and untrained limbs was greatest in subjects with the highest rate-pressure product at angina before training. Control subjects showed no change in any exercise measurement.Exercise training increases the exercise capacity of untrained limbs in patients with angina pectoris by a generalized training effect not dependent on adaptations In trained skeletal muscle. The improvement for both trained and untrained limbs results from a reduced rate-pressure product at subanginal work loads rather than from an increase in myocardial oxygen delivery. Subjects with the highest pretraining coronary arterial oxygen supply at the onset of angina benefit most from physical training.