We studied 19 patients with angiographically proven coronary artery disease and left ventricular dysfunction (ejection fraction 33% +/- 8%) by resting technetium-99m methoxyisobutylisonitrile (99mTc-MIBI) and rest-redistribution thallium-201 cardiac imaging. Thallium and 99mTc-MIBI studies were visually analysed. Of 285 segments, 203 (71%) had normal thallium uptake, 48 (17%) showed reversible thallium defects and 34 (12%) showed irreversible thallium defects. Of these 34 irreversible thallium defects, 19 (56%) were moderate and 15 (44%) were severe. Of the corresponding 285 segments, 200 (70%) had normal 99mTc-MIBI uptake, while 37 (13%) showed moderate and 48 (17%) showed severe reduction of 99mTc-MIBI uptake. Myocardial segmental agreement for regional uptake score between initial thallium and resting 99mTc-MIBI images was 90% (kappa = 0.78). Segmental agreement between delayed thallium and resting 99mTc-MIBI images was 77% (kappa = 0.44). In particular, in 26 (9%) segments 99mTc-MIBI uptake was severely reduced while delayed thallium uptake was normal or only moderately reduced. These data suggest that although rest-redistribution thallium and resting 99mTc-MIBI cardiac imaging provide concordant results in the majority of myocardial segments, some segments with severely reduced resting 99mTc-MIBI uptake may contain viable but hypoperfused myocardium. Thus, conclusions on myocardial viability based on 99mTc-MIBI uptake should be made with caution in chronic coronary artery disease.