Emergency department chest pain centers (CPCs) vary in their approach to patients with chest pain and nonischemic electrocardiograms (ECG). Although single-photon emission computed tomography (SPECT) myocardial perfusion imaging has been evaluated in this setting, both acutely at rest and after stress, we questioned its application in all patients. We prospectively evaluated the utility of selective SPECT imaging in a CPC (i.e., rest SPECT for ongoing pain, stress SPECT if unable to undergo exercise electrocardiography) and its impact on the overall disposition of all emergency department chest pain patients. Over 3 years, 2,601 patients were evaluated in a CPC (2,211 [85%] were sent home, 390 [15%] were hospitalized). Of 390 CPC patients hospitalized, 182 (47%) were diagnosed with coronary artery disease at the time of hospital discharge. Only 28 patients (1.1%) had an acute myocardial infarction. After 3 years, the proportion of all chest pain patients hospitalized and those diagnosed as "rule-out myocardial infarction" decreased from 53% to 41% and 32% to 18% of all chest pain patients, respectively (both p <0.0001). Overall, 906 patients (35%) required SPECT imaging to complete the CPC evaluation. Had SPECT imaging not been performed selectively, and all 906 patients been admitted, 762 (29%) would have been hospitalized unnecessarily based on the final diagnoses. Alternatively, sending all these patients home would have resulted in 144 (6%) inappropriate discharges of patients with coronary artery disease. A CPC protocol using the selective use of SPECT imaging permits the complete evaluation of all patients in the CPC, significantly reduces hospitalizations for chest pain, and restricts hospital admission to more appropriate patients.