We examined the relations among body mass index (BMI), extent of coronary artery disease, and frequency of high-risk coronary anatomy (HRCA) in 928 consecutive patients who underwent coronary angiography during a 6-month period. HRCA was defined as >or=50% stenosis of the left main coronary artery and/or significant 3-vessel coronary artery disease (>or=70% narrowing). BMI was classified into 5 subgroups: low (<21 kg/m(2)), normal (21 to 24 kg/m(2)), overweight (25 to 29 kg/m(2)), obese (30 to 34 kg/m(2)), and severely obese (>or=35 kg/m(2)). Obese patients (BMI >or=30 kg/m(2)) were younger (61.4 +/- 10.7 vs 65.3 +/- 11.4 years, p <0.0001) and had higher prevalences of hyperlipidemia, systemic hypertension, and diabetes mellitus. HRCA was present less often in obese patients (56 of 245, 23%, vs 250 of 683, 37%, p = 0.0002). Multivariate regression analysis showed that advancing age (p <0.0001), male gender (p = 0.007), diabetes mellitus (p = 0.0004), and hyperlipidemia (p = 0.0008) were independent predictors of high-risk anatomy, whereas obesity remained a significant negative independent predictor (p = 0.02). Late (30 to 36 months) mortality was not different between obese (6.9%) and nonobese (8.2%) patients but was significantly higher in patients with HRCA (12.4%) than in those without HRCA (5.6%, p = 0.0003). In conclusion, obese patients who were referred for coronary angiography were younger and had a lower prevalence of HRCA. Obese patients were probably referred for angiography at an earlier stage of their disease, thus explaining the "obesity paradox" in several reports of better short-term outcome in obese patients who undergo cardiac procedures.