Dobutamine echocardiography (DE), magnetic resonance imaging (MRI), and thallium redistribution (TS) are used to assess cardiac viability. However, these modalities sometimes yield contradictory results. Our aim was to establish the degrees of agreement among DE, MRI, and TS in identifying myocardial viability and to analyze the minimum critical mass of live (viable) cells required for each test to identify viability. A prospective study was done in which DE, MRI, and TS were consecutively performed in 10 ischemic patients scheduled for heart transplantation. The explanted heart was analyzed to quantify the amount of live cells per segment. The pathologic data were compared with the test results to analyze the minimum mass of viable cells required by each technique to identify viability. Mean age was 58 +/- 8 years (8 men). The mean ejection fraction was 0.27 +/- 0.04. Seven patients had severe cardiac failure (New York Heart Association functional class IV) and 6 patients had refractory angina. A total of 150 cardiac segments were analyzed. Among the 150 segments, 107 (71.3%) showed some degree of myocardial necrosis. Mean total area, mean fatty area, and mean necrotic area per segment were 2.53 +/- 0.7, 0.13 +/- 0.2, and 0.55+/-0.5 cm(2), respectively. As expected, a higher amount of necrotic tissue was found in nonviable segments. From the 150 segments, DE identified 90 as viable and 60 as nonviable. These data were similar to that of MRI (98 viable and 52 nonviable). A higher proportion of viable segments was found by TS (117 viable vs 33 nonviable). The concordance between DE and TS was only moderate (kappa 0.49). The agreement between MRI and TS also showed moderate concordance (kappa 0.56). The highest agreement was found between DE and MRI (kappa 0.73). Thus, discrepancies in assessing viability by DE, MRI, and TS may be due to differences in the minimum critical mass of live myocytes required by each technique to diagnose viability. Thallium requires a lesser amount of live tissue than DE or MRI to detect viability; also, its maximum diagnostic efficiency is obtained with lesser amounts of live tissue on each segment. These considerations should be taken into account when these diagnostic tests are used for the detection of viability before revascularization procedures.