An exercise test based on arm work was evaluated in a series of 33 male patients, mean age 52 years, with ischemic heart disease. The responses to arm exercise on a modified table-mounted bicycle ergometer and to standard bicycle exercise were compared. Twenty six of 33 patients (79 per cent) had identical end-points with both tests. Three patients had an ischemic response, i.e., significant ST abnormality and/or angina pectoris during leg work only, and four patients during arm work only. 41 per cent of the peakload during leg exercise. Mean values were 181 and 439 kpm./min. (p less than 0.001). Comparison of individual data on peak load demonstrated only a weak correlation between arm and leg work capacity (r = 0.37, p less than 0.05). Peak heart rate was slightly higher during leg work, 129 compared to 122 beats/min. (p less than 0.05) but the mean heart rate-systolic blood pressure products were not significantly different. A subgroup of seven patients had a history of angina pectoris preferentially precipitated by arm work but their physiological responses did not differ significantly from those of patients without a history of arm work sensitivity. The data indicate that arm work is a satisfactory alternate diagnostic test method with respect to myocardial ischemia, but measurements of physical work capacity defined as aerobic capacity, cannot be based on arm work.