Hypoadiponectemia was observed to correlate with the prevalence or extent of coronary artery disease (CAD), the relationship between metabolic syndrome and fasting serum adiponectin concentration in CAD patients are not well elucidated. Fasting blood samples were obtained from 98 CAD patients. Metabolic syndrome and its components were defined using the diagnostic criteria of the International Diabetes Federation. Adiponectin concentrations were measured using a commercial enzyme immunoassay kit. Fifty patients with CAD (51.0%) had metabolic syndrome. For this group of patients, fasting adiponectin concentrations were found to correlate inversely with metabolic syndrome (p=0.009). Fasting adiponectin values for these subjects also tended to decrease as the number of diagnostic criteria for metabolic syndrome increased (p=0.024). CAD patients with hyperlipidemia (p=0.002), obesity (p=0.030) or receiving statin therapy (p=0.005) had lower serum adiponectin values. By univariate linear regression analysis, fasting serum adiponectin values were positively correlated with age (r=0.242; p=0.017) and high density lipoprotein-cholesterol concentration (HDL-cholesterol; r=0.267; p=0.008) but were negatively correlated with triglyceride concentration (TG; r=-0.251; p=0.013). Multivariate forward stepwise linear regression analysis of the significant variables revealed that HDL-cholesterol concentration (R square=0.071, p=0.008) and age (R square=0.039, p=0.044) are the independent predictors of fasting serum adiponectin concentration for patients with CAD. Serum adiponectin concentration is inversely correlated with metabolic syndrome and the number of metabolic syndrome criteria in patients with CAD. For these patients HDL-cholesterol concentration and age are independent predictors of the serum adiponectin value.