Relationships between enzyme estimates (EE) and histologic measurements of infarct size (HIS) were analyzed in 14 conscious dogs with permanent occlusion of the circumflex coronary artery. EE were derived from serial CPK, CPK-MM, CPK-MB, and CPK-BB. Estimates were obtained using methods of Shell et al., Norris et al., and Roberts et al. HIS was determined from multiple histologic sections 5-6 days after infarction. In 14 animals, HIS ranged from 0.1-26.6 grams. Regression analysis demonstrated poor correlation (r2 less than 0.06) between EE by each method and HIS. Using the Shell method and restricting the analysis to HIS of less than 13 grams, improved the correlation (r2 = 0.42). Correlation in this subgroup could be further improved by using CPK-MM and MB data, elimination of animals with high background values, or limiting analysis to the portion of the curve where LDH isoenzymes indicated a cardiac zone. When the analysis was restricted to animals with HIS of less than 13 grams, Norris and Roberts modifications as compared to Shell's method improved correlation with CPK-MB but reduced correlation using CPK and CPK-MM data. In this subgroup, each method overestimated extent of HIS; Shell greater than Norris greater than Roberts. In this study it was not possible to distinguish small from large HIS by serum enzyme measurements.