We investigated the association among plasma levels of oxidatively modified low-density lipoprotein (Ox-LDL), high-density lipoprotein-cholesterol (HDL-C), and the prevalence of coronary artery disease (CAD) in a case-control study. Cases (n = 183, male [M]/female [F]:138/45, age: 64.9 +/- 10.6 years) were defined as patients with angiographically proven coronary atherosclerosis (>/=50% stenosis) and controls were subjects with normal coronary arteries (n = 74, M/F:36/38, age: 57.6 +/- 14.4 years). Plasma Ox-LDL levels were measured by a sensitive detection method using the monoclonal antibody DLH3. In women, both Ox-LDL and lipid variables were similar between cases and controls. In men, cases had significantly lower (P <.05) levels of HDL-C (39.1 +/- 10.3 v 42.8 +/- 10.9 mg/dL) and apolipoprotein (apo) A-I than controls, while the difference in Ox-LDL between cases and controls was not significant (1.05 +/- 0.79 and 0.83 +/- 0.65 ng/10 microg LDL protein, respectively). However, HDL-C levels interacted with the association between Ox-LDL levels and CAD in males: increased Ox-LDL levels were significantly associated with CAD after controlling for age when HDL-C levels were high, but were not associated with CAD when HDL-C levels were low, as assessed by a multiple logistic regression analysis. In addition, the combination of HDL-C and Ox-LDL levels was a better indicator for CAD in males than HDL-C levels alone (-2 log likelihood, 24.1 v 19.4) after controlling for age and conventional risk factors of CAD, while Ox-LDL levels were not significantly associated with CAD. HDL-C levels interact with the association between plasma Ox-LDL levels and CAD in men, and increased Ox-LDL levels are an indicator of CAD in male subjects with high HDL-C levels.