Pulmonary embolism is a serious and potentially fatal disorder. Pulmonary embolism risk stratification may allow early hospital discharge and outpatient treatment for low-risk patients. Also, it may prevent death by early medical intervention in high-risk groups. We evaluated objectively confirmed pulmonary embolism in 126 patients by multidetector computed tomographic pulmonary angiography at a single center from January 2008 to January 2010. The Pulmonary Severity Embolism Index (PESI), the right ventricle (RV) to left ventricle (LV) diameter (RV/LV) ratio and the vascular obstruction index (VOI) were derived from data extracted from electronic hospital records and image database. A total of six out of 96 patients (6.3%) died during follow-up. There was an association between PESI and mortality (P-value < 0.001 χ² test). PESI class I-II had a 100% negative predictive value for death in 90 days. No association was found between the RV/LV ratio, the VOI and mortality (P-value > 0.05 χ² test). Also, no association was found between the RV/LV ratio and the VOI and PESI (P-value > 0.05 χ² test). PESI is an accurate tool for pulmonary embolism prognostic stratification. It safely discriminates low-risk from high-risk patients regarding death outcome. We were unable to demonstrate an association between image scores and mortality.