OBJECTIVES: To determine the incidence of clinical pulmonary embolism (PE) in a population with severe congestive heart failure (CHF) admitted to a coronary care unit (CCU), and to identify clinical predictors of PE in this population. DESIGN AND SETTING: Prospective, observational study performed in a CCU of a tertiary care hospital between July 2001 and March 2003. PATIENTS: One hundred ninety-eight patients with severe decompensated CHF. MEASUREMENTS AND RESULTS: Of 198 patients recruited, 18 patients (9.1%) received a diagnosis of PE during their hospitalization. Deep vein thrombosis was demonstrated in 8 of 18 patients (44.4%) with PE. Thromboprophylaxis was used by 12 of 18 patients (66.7%) with PE and 126 of 180 patients (70%) without PE (p = 0.77). Both groups were similar with respect to mean age (68.2 +/- 14.1 years vs 69.6 +/- 13.4 years [+/- SD]), proportion of male patients (61.1% vs 55.1%), markers of CHF severity (New York Heart Association functional class > II, ejection fraction < 30%, Na < 136 mEq/L, ischemic etiology), and comorbid conditions (diabetes mellitus, atrial fibrillation, chronic renal failure, hypertension) [p = not significant]. The presence of PE was significantly associated with cancer (relative risk [RR], 8.4; 95% confidence interval [CI], 3.9 to 18.1), immobilization (RR, 5.4; 95% CI, 2.0 to 14.4), previous venous thromboembolism (VTE) [RR, 4.4; 95% CI, 1.7 to 11.3], COPD (RR, 3.1; 95% CI, 1.03 to 9.2), and right ventricle (RV) abnormality (RR, 3.3; 95% CI, 1.3 to 8.0). In a multiple logistic regression analysis, only cancer (odds ratio [OR], 26.9; 95% CI, 4.9 to 146.8), RV abnormality (OR, 9.7; 95% CI, 2.2 to 42.6), and previous VTE (OR, 9.1; 95% CI, 1.28 to 64.7) remained independently associated with PE. CONCLUSIONS: In patients with severe decompensated CHF admitted to a CCU, the incidence of clinical PE is very high despite adequate prophylaxis. Traditional risk factors seemed to play an important role in determining the risk of PE in this population.