We have investigated the role of muscle mass, natriuretic peptides and adipokines in explaining the obesity paradox. Background: The obesity paradox relates to the association between obesity and increased survival in patients with coronary heart disease (CHD) or heart failure (HF). Methods: Prospective study of 4046 men aged 60–79 years followed up for a mean period of 11 years, during which 1340 deaths occurred. The men were divided according to the presence of doctor diagnosed CHD and HF: (i) no CHD or HF ii), with CHD (no HF) and (iii) with HF. Results: Overweight (BMI 25–9.9 kg/m2) and obesity (BMI ≥ 30 kg/m2) were associated with lower mortality risk compared to men with normal weight (BMI 18.5–24.9 kg/m2) in those with CHD [hazards ratio (HR) 0.71 (0.56,0.91) and 0.77 (0.57,1.04); p = 0.04 for trend] and in those with HF [HR 0.57 (0.28,1.16) and 0.41 (0.16,1.09; p = 0.04 for trend). Adjustment for muscle mass and NT-proBNP attenuated the inverse association in those with CHD (no HF) [HR 0.78 (0.61,1.01) and 0.96 (0.68,1.36) p = 0.60 for trend) but made minor differences to those with HF [p = 0.05]. Leptin related positively to mortality in men without HF but inversely to mortality in those with HF; adjustment for leptin abolished the BMI mortality association in men with HF [HR 0.82 (0.31,2.20) and 0.99 (0.27,3.71); p = 0.98 for trend]. Conclusion: The lower mortality risk associated with excess weight in men with CHD without HF may be due to higher muscle mass. In men with HF, leptin (possibly reflecting cachexia) explain the inverse association.