Decision making and risk stratification for patients with acute chest pain, nondiagnostic electrocardiogram results, and normal troponin levels are challenging. The aim of this study was to optimize the clinical history for the evaluation of these patients. A total of 1,011 patients presenting to an emergency department were included. The following data were collected: clinical presentation (pain characteristics and number of pain episodes), coronary risk factors, previous ischemic heart disease, and extracardiac vascular disease (peripheral artery disease, stroke, or creatinine >1.4 mg/dl). Two different predictive models were calculated according to the end points: model 1 for 1-year major events (death or myocardial infarction) and model 2 for 30-day cardiac events (major events or revascularization). For 1-year major events, model 1 showed optimal discrimination capacity (C statistic = 0.80), which was significantly better than that of model 2 (C statistic = 0.77, p = 0.04). With respect to 30-day cardiac events, however, discrimination was lower in the 2 models, without differences between them (C statistic = 0.74 vs 0.75, p = NS). Using model 1, a large low-risk subgroup with <3 predictive variables could be defined, including 442 patients (44% of the total population) with a 1.4% rate of 1-year major events; however, the incidence of 30-day cardiac events (8%) was not negligible, mainly because of revascularizations. In conclusion, in patients with acute chest pain of uncertain coronary origin, clinical predictive models afford good risk stratification for long-term major events. Short-term outcomes, including revascularization, however, are not predicted as well. Therefore, ancillary tools, such as noninvasive stress tests, should be implemented for decision making at initial hospitalization or discharge.