In ischaemic heart disease patients, transient left ventricular dysfunction is observed due to post-exercise stunning. The aim of this study was to determine whether transient left ventricular dysfunction could also be seen after short-acting pharmacological stress (adenosine triphosphate). A 1 day rest/stress gated myocardial single photon emission computed tomography was performed on 362 patients suspected of having ischaemic heart disease by exercise (n=199) or short-acting pharmacological stress (n=163). Left ventricular ejection fraction were estimated both at rest and stress. Based on perfusion findings, patients were subdivided into ischaemia, fixed defect and normal group. For the ischaemia and fixed defect group, left ventricular ejection fraction after stress was significantly decreased compared with the resting value by exercise stress (ischaemia group, 57.5+/-11.0 vs 60.4+/-10.4; fixed defect group, 47.7+/-16.7 vs 49.6+/-16.8; P<0.01), but not by pharmacological stress (ischaemia group, 55.8+/-13.4 vs 57.1+/-13.8; fixed defect group, 50.8+/-13.5 vs 50.6+/-13.1; P=NS). In the normal group, left ventricular ejection fraction after stress was not significantly changed by either exercise (65.7+/-10.4 vs 66.8+/-10.2; P=NS) or pharmacological stress (63.0+/-11.7 vs 64.0+/-12.1; P=NS). It is concluded that a transient decrease in left ventricular ejection fraction after stress was observed following post-exercise, not following a short-acting pharmacological stress in patients showing perfusion abnormalities. Transient left ventricular dysfunction may be the result of post-exercise stunning, not from subendocardial hypoperfusion induced by short-acting pharmacological stress.