OBJECTIVE:To compare the ability of dobutamine and exercise stress to induce myocardial ischaemia and perfusion heterogeneity under routine clinical circumstances. DESIGN:86 active patients without previous myocardial infarction were studied by dobutamine and exercise stress protocols and coronary angiography. During both tests patients underwent electrocardiography, digitised echocardiography, and perfusion scintigraphy using Tc-99m methoxybutylisonitrile (MIBI) single photon emission computed tomography. MAIN OUTCOME MEASURE:Coronary disease defined as an ST segment depression of > or = 0.1 mV, a resting or stress induced perfusion defect, or a resting or stress induced wall motion abnormality on exercise and dobutamine stress testing. RESULTS:Dobutamine stress was submaximal in 51 patients because of ingestion of beta adrenoceptor blocking agents on the day of the test (n = 25) or failure to attain the peak dose owing to side effects (n = 28). Exercise was limited in 23 patients by non-cardiac symptoms. The peak heart rate with dobutamine was less than that attained with exercise (105 (25) v 132 (24) beats/min, P < 0.0001); the response to maximal dobutamine stress significantly exceeded that to submaximal stress. Peak blood pressure was greatest with exercise (206 (27) v 173 (25) mm Hg, P < 0.001), values at maximal and submaximal dobutamine stress being comparable. Electrocardiographic evidence of ischaemia was induced less frequently by dobutamine than exercise (32% v 77% of the 56 patients with significant coronary disease, P < 0.01), as was abnormal wall motion (54% v 88%, P < 0.001). Ischaemia was induced more readily with maximal stress of either type; thus the sensitivities of dobutamine and exercise echocardiography were comparable only in patients undergoing a maximal dobutamine testing (73% v 77%, NS). Perfusion heterogeneity was induced in 58% of patients with coronary disease at submaximal dobutamine stress, 73% at maximal dobutamine stress, and 73% at exercise stress (NS). Among 30 patients without coronary stenoses, normal function was obtained in 83% of echocardiography studies with dobutamine and in 80% with exercise (NS). Normal perfusion was identified in 70% of these patients at exercise MIBI, and 68% at dobutamine stress (NS). CONCLUSIONS:In a group of patients studied under normal clinical circumstances antianginal treatment and inability to complete the stress protocol are frequent and compromise the capacity of dobutamine stress to induce ischaemia. In contrast, the induction of perfusion heterogeneity is less susceptible to submaximal stress.