To determine the contractile reserve of the left ventricle during reperfusion as a predictor of myocardial viability in patients undergoing coronary artery bypass graft surgery, we measured the response of left ventricular regional wall motion and thickening by using dobutamine stress echocardiography (DSE) after myocardial revascularization. All patients were monitored with radial and pulmonary arterial catheters, transesophageal echocardiography, standard five-lead clinical electrocardiography, and three-channel Holter electrocardiography. Immediately after separation from cardiopulmonary bypass, dobutamine was administered IV starting at 5 microg. kg(-1). min(-1), with increases in rate every 3 min to 10, 20, 30, and 40 microg. kg(-1). min(-1). Within 1 wk after surgery, resting and redistribution thallium-201 myocardial perfusion imaging (thallium studies) was performed to assess the relationship between the intraoperative contractile response and myocardial viability. One-hundred patients completed DSE up to 10 microg. kg(-1). min(-1), and 85 patients received the larger escalating doses of the DSE. Seventy-two patients had postoperative thallium studies. At the completion of the small-dose dobutamine infusion, 689 (97.7%) of 705 segments had a normal response (improvement), and 16 segments (2.3%) had a positive response (deterioration). During large-dose dobutamine infusion, 577 (95.8%) of 602 segments had a normal response, and 25 segments (4.2%) had a positive response. Myocardial segments that had a positive response during large-dose DSE (48%) were more likely to be considered as nonviable on postoperative thallium studies compared with segments that had a normal response (14.7%) (P < 0.00001). By using thallium studies as the reference standard, the sensitivity of DSE was low (31% and 48% for small- and large-dose DSE, respectively) in predicting nonviable myocardium. However, the specificity was higher (86% and 85% for small- and large-dose DSE, respectively). In a separate analysis of patients who developed new regional wall motion abnormalities (RWMA) in the early intraoperative reperfusion period, 15 (75%) of 20 abnormally contracting myocardial segments had normal postoperative thallium studies. Our results demonstrate that a normal response to DSE is highly specific for viable myocardium; however, a positive response to DSE has low sensitivity in predicting nonviable myocardium. The majority of new postbypass regional wall motion abnormalities appear to be related to stunned myocardium.