Effective use of creatine kinase M and B isoform (CK-MB) test results for the diagnosis or exclusion of acute myocardial infarction (AMI) requires repeated analyses at defined intervals after the onset of chest pain. An increase in serum CK-MB is evaluated relative to typical angina-like chest pain of at least 30 minutes duration and electrocardiographic changes per World Health Organization (WHO) standards. Two or more of the three findings must be abnormal for the diagnosis of AMI. We assessed the appropriateness (clinical usefulness) and cost-effectiveness of CK-MB mass and activity levels in the care of patients with complaints of chest pain seen in our emergency department. The total CK and CK-MB (mass and activity) results for 860 patients during 2 months in 1995 were reviewed retrospectively relative to the discharge diagnosis. According to the discharge diagnoses based on the WHO diagnostic criteria for AMI, only 54 patients (6.3%) had acute myocardial infarction. Three or more CK-MB tests (mass and activity measurements) were performed in 60.5% of patients, and only one test was performed in the remaining 39.5%. Abnormal CKMB mass or activity was determined with the following criteria: Protocol 1: CK-MB mass >5 ng/mL, total CK >194 U/L, index >2.5%; Protocol 2: CK-MB activity >16 U/L, total CK >65 U/L, index >5%; Protocol 3: CK-MB activity >10 U/L, total CK >65 U/L, index >2.2%. Based on these criteria, CK-MB mass and activity protocols gave positive predictive values of 38.2%, 36.1%, and 35.4% for protocols 1, 2, and 3, respectively. The negative predictive value was 99% for CKMB mass (Protocol 1), and 98% and 98.9% for CK-MB activity (Protocols 2 and 3). These results indicate that both CK-MB mass and activity protocols have equivalent clinical usefulness. However, given the technologic limitation of two platforms for Protocol 1 vs one assay platform for Protocols 2 and 3, the CK-MB activity protocol is more cost-effective for patient testing in the emergency department laboratory.