High-density lipoprotein cholesterol (HDL) and physical fitness (PF) have both been shown to predict cardiovascular disease (CVD), particularly coronary heart disease (CHD). Increased PF is associated with increased HDL and may partly explain the benefit of HDL. We tested the hypothesis that PF influences the prognostic impact of HDL for CHD and also for CHD-, CVD- and all-cause death.HDL was measured 1979–1982 in 1357 healthy men aged 44–69 years followed up to 28 years. PF was measured using bicycle exercise test. Hazard ratios (HRs) adjusted for age, smoking, systolic blood pressure, and total cholesterol and further for PF between HDL quartiles were calculated using Cox proportional survival model.The highest HDL quartile was associated with lower risk of CHD (HR: 0.57, 95% confidence interval [CI]: 0.43–0.74), fatal CHD (HR: 0.56, CI: 0.36–0.86), fatal CVD (HR: 0.64, CI: 0.46–0.88) and all-cause death (HR: 0.80, CI: 0.65–0.99) compared to the lowest quartile. Adjustments for PF or changes in PF over 8.6 years did not change the results except for all-cause death, which was not significantly different between HDL quartiles. We found no interaction between HDL and PF.HDL is a strong predictor of long term risk of CHD, fatal CHD and fatal CVD in healthy middle-aged men. Physical fitness or its changes had no impact on the ability of HDL to predict CHD.