BACKGROUND Newer diagnostic modalities such as serum markers and acute rest myocardial perfusion imaging (MPI) have been evaluated diagnostically in patients with chest pain in the emergency department (ED), but never concurrently. We compared these two modalities in distinguishing patients in the ED with symptomatic myocardial ischemia from those with non-cardiac causes. METHODS Serum markers and acute technetium-99m sestamibi/tetrofosmin rest MPI were obtained in 75 patients admitted to the ED with chest pain and nondiagnostic electrocardiograms. Venous samples were drawn at admission and 8 to 24 hours later for total creatine kinase, CK-MB fraction, troponin T, troponin I, and myoglobin. Three nuclear cardiologists performed blinded image interpretation. Coronary artery disease (CAD) was confirmed either by diagnostic testing or by the occurrence of myocardial infarction (MI). RESULTS Acute rest MPI results were abnormal in all 9 patients with MI. An additional 26 patients had objective evidence of CAD confirmed by diagnostic testing. The sensitivity of acute rest MPI for objective evidence of CAD was 73%. Serum troponin T and troponin I were highly specific for acute MI but had low sensitivity at presentation. Individual serum markers had very low sensitivity for symptomatic myocardial ischemia alone. In the multivariate regression model, only acute rest MPI and diabetes were independently predictive of CAD. CONCLUSIONS At the time of presentation and 8 to 24 hours later, acute rest MPI has a better sensitivity and similar specificity for patients with objective evidence of CAD when compared with serum markers.