© 2018 Patients admitted with suspected acute coronary syndrome (ACS) in whom the diagnosis is not confirmed are poorly characterized. In a contemporary registry of consecutive patients hospitalized with suspected ACS as the primary diagnosis, we assessed characteristics on admission and in-hospital and 6-month mortality of patients discharged with other diagnoses and compared this subgroup with true ACS patients. Of 2557 patients included, 9.0% were discharged with a non-ACS diagnosis such as nonspecific chest pain, myopericarditis, stress cardiomyopathy, hemodynamic disturbances, heart failure, myocardial, pulmonary or valvular disease, or others. Compared with true ACS patients, those with other diagnoses were younger, more often female, and had less cardiovascular risk factors. Both groups had comparable rates of nonchest pain presentation and similar hemodynamic characteristics on admission. Non-ACS patients presented less often with Q waves or with ST-segment or T-wave changes and had a lower Global Registry of Acute Coronary Events score than true ACS patients. In-hospital (4.3 vs 4.0%, respectively, p = 0.834) and 6-month (5.4 vs 8.0%, respectively, p = 0.163) mortality rates were comparable in both groups. However, if patients in the non-ACS group were divided into subgroups with nonspecific chest pain (6.2% of total) or other diagnoses (2.8% of total), major differences in in-hospital (0.0 vs 13.9%, respectively, p < 0.001) and 6-month (0.7 vs 15.7%, respectively, p < 0.001) mortality rates would become apparent and remain after multivariable adjustment. In conclusion, in a non-negligible proportion of patients hospitalized with suspected ACS, this diagnosis is not confirmed. Prognosis of these patients follows a bimodal pattern, being excellent in those with nonspecific chest pain but worse than that of true ACS patients in the rest. Efforts are necessary to ensure prompt identification and early risk stratification of these patients allowing appropriate management decisions.