BACKGROUND Echo-dipyridamole test is an useful tool for non-invasive demonstration of inducible myocardial ischemia in patients with coronary artery disease, its sensitivity being consistently higher as compared with classic exercise stress testing. However, in patients with single vessel who often perform a normal or borderline stress test, even the sensibility of echo-dipyridamole test is comparatively reduced. METHODS In 19 patients with clinically suspected angina (effort-related in 4, at rest in 8, mixed in 7) and normal exercise stress test, standard echo-dipyridamole test was performed. Thereafter, rapid atrial pacing, a procedure associated with a sharp increase of myocardial oxygen consumption, was performed by means of a transoesophageal catheter during the proceeding 5 min and during 4 min of repeated dipyridamole 0.56 mg/kg infusion. RESULTS Standard echo-dipyridamole test induced ventricular wall motion abnormalities in 3 patients (one with borderline exercise stress test), whereas repeated pacing-sensitized procedure obtained wall motion abnormalities (apical, septal and lateral) in the same and in 4 additional patients. Coronary angiography demonstrated > 70% stenosis in 8/19 patients (single vessel disease in 5), 7 of whom had been correctly recognized by pacing-dipyridamole test; therefore, sensitivity of the latter as compared with standard dipyridamole test was 87% and 37% respectively. The anatomic correlation of induced wall motion abnormalities with coronary arterial stenosis was demonstrated in all cases. Specificity was 100% with both methods. CONCLUSIONS Although limited by its restricted patient population, this study suggests that atrial pacing, performed via transoesophageal catheter, can significantly improve the positive predictive value of echo-dipyridamole test in coronary artery disease. By means of this simple procedure, the possibility of non invasively diagnosing even single vessel stenosis in patients with inconclusive exercise stress testing can be significantly improved.