Exercise stress testing is routinely used for the noninvasive assessment of coronary artery disease and is considered a safe procedure. However, the provocation of severe ischemia might potentially cause delayed recovery of myocardial function. To investigate the possibility that maximal exercise testing could induce prolonged impairment of left ventricular function, 15 patients with angiographically proved coronary disease and 9 age-matched control subjects with atypical chest pain and normal coronary arteries were studied. Radionuclide ventriculography was performed at rest, at peak exercise, during recovery and 2 and 7 days after exercise. Ejection fraction, peak filling and peak emptying rates and left ventricular wall motion were analyzed. All control subjects had a normal exercise test at maximal work loads and improved left ventricular function on exercise. Patients developed 1 mm ST depression at 217 +/- 161 s at a work load of 70 +/- 30 W and a rate-pressure product of 18,530 +/- 4,465 mm Hg x beats/min. Although exercise was discontinued when angina or equivalent symptoms occurred, in all patients diagnostic ST depression (greater than or equal to 1 mm) developed much earlier than symptoms. Predictably, at peak exercise patients showed a decrease in ejection fraction and peak emptying and filling rates. Ejection fraction and peak emptying rate normalized within the recovery period, whereas peak filling rate remained depressed throughout recovery (p less than 0.002) and was still reduced 2 days after exercise (p less than 0.02). In conclusion, in patients with severe impairement of coronary flow reserve, maximal exercise may cause sustained impairement of diastolic function.(ABSTRACT TRUNCATED AT 250 WORDS)