OBJECTIVE: Vasodilator stress echocardiography can cause myocardial ischaemia in patients with severe aortic valve stenosis and angiographically normal coronary arteries. The aim of the study was to determine the mechanism of ischaemia in this clinical model. METHODS: The study group comprised patients with severe aortic valve stenosis and normal coronary arteries: 25 patients (17 males, eight females; age 63 (SD 11) years) underwent a high dose (up to 0.84 mg/kg over 10 min) dipyridamole echocardiography test both before (2-4 d) and after (10-15 d) aortic valve replacement. Mean aortic pressure gradient was 96 (15) mm Hg, with a left ventricular mass index of 228 (49) g/m2. The dipyridamole echocardiography test was well tolerated and interpretable in all patients. RESULTS: Dipyridamole infusion induced chest pain in seven patients before and in no patient after surgery (28 v 0%, P < 0.01), ST segment depression in 12 patients before and two after surgery (48 v 8%, P < 0.01), and a transient regional dyssynergy in 10 patients before and two after surgery (40 v 8%, P < 0.01). In the preoperative evaluation, patients with an echocardiographically positive dipyridamole echocardiography test were comparable with patients with negative test as far as left ventricular mass index [240 (67) v 230 (64) g/m2, NS] and mean aortic pressure gradient [95 (22) v 92 (21) mm Hg, NS] were concerned. When compared to the preoperative assessment, the resting echo assessment in the postoperative evaluation showed unchanged values of left ventricular mass index [pre 228 (49) g/m2 v post 220 (36) g/m2, NS], but markedly decreased values of mean aortic gradient [pre 95 (15) mm Hg v post 22 (5) mm Hg, P < 0.01] and left ventricular wall stress index [pre 134 (30) g/cm2 v post 89 (19) g/cm2]. CONCLUSIONS: Dipyridamole echocardiography is a suitable clinical technique for assessing the ischaemic vulnerability of the left ventricle in severe aortic valve stenosis with angiographically normal coronary arteries. The frequent disappearance of the ischaemic response early after aortic valve replacement suggests that haemodynamic factors such as compressive diastolic wall stress or afterload reduction are important components of myocardial ischaemic vulnerability under these circumstances.