Recently, the use of 18F-fluorodeoxyglucose (FDG) with SPECT has been proposed to identify viable myocardium. Thallium-201 rest-redistribution SPECT is frequently used for this purpose. The aim of this study was to compare the relative merits of the two techniques to predict improvement of regional left ventricular function after revascularization. Methods: Twenty- four patients scheduled for revascularization were studied. Regional contractile function was evaluated with echocardiography before and 3 mo after revascularization. The patients underwent 201Tl rest-redistribution SPECT and FDG SPECT on the same day; the early 201Tl image obtained in the rest-redistribution protocol was also used as a perfusion study for comparison with FDG SPECT. The SPECT data were analyzed semiquantitatively using polar maps. For analysis of echo and SPECT data, a 13-segment model was used. Dysfunctional segments were classified as viable on FDG SPECT if normal perfusion or mismatch pattern (increased FDG uptake in perfusion defects) was seen. For 201Tl rest-redistribution SPECT, criteria for viability included the percentage of 201Tl uptake and reversibility of defects. Results: On echocardiography, 106 segments were dysfunctional Recovery of function was observed in 36 segments, whereas function did not improve in 70. FDG SPECT showed a sensitivity of 89% [95% confidence interval (CI) 79%-99%] and a specificity of 81% (95% CI 72%-90%). Thallium-201 rest-redistribution SPECT had a sensitivity of 67% (95% CI 52%-82%) and a specificity of 77% (95% CI 67%-87%). Conclusion: Our data show that the diagnostic accuracies (to predict improvement of contractile function after revascularization) of both techniques were not significantly different, indicating that both imaging modalities may provide comparable information concerning prediction of improvement of function.