Vascular endothelial dysfunction has been demonstrated in overweight or obese patients, but the molecular basis for this link has not been clarified. We asked what the relationship was between adiponectin, an adipose-specific molecule, and endothelial function. Forearm blood flow (FBF) was measured during reactive hyperemia by using strain-gauge plethysmography in 76 Japanese subjects without a history of cardiovascular or cerebrovascular disease, diabetes mellitus, hepatic, or renal disease. The peak FBF and total reactive hyperemic flow [flow debt repayment (FDR)] during reactive hyperemia were correlated with waist circumference (r = -0.418 and -0.414, respectively) and body mass index (r = -0.597 and -0.626, respectively). After correcting for age, gender, and body mass index, the peak FBF was correlated with systolic blood pressure (r = -0.294; P = 0.010), free fatty acid (FFA) (r = -0.331; P = 0.004), and adiponectin in log 10 (r = 0.492; P < 0.001), and FDR was correlated with adiponectin in log 10 (r = 0.462; P = 0.001). In stepwise multiple regression analyses, predictive variables for peak FBF were adiponectin in log 10 (r = 0.468) and FFA (r = -0.292; r(2) = 0.487; P < 0.0001); and predictive variables for FDR were adiponectin in log 10 (r = 0.474) and FFA (r = -0.275; r(2) = 0.346, P < 0.0001). Endothelial function was impaired in proportion to the severity of obesity, and the level of severity was closely related to plasma adiponectin levels. Adiponectin may play a protective role against the atherosclerotic vascular change, and loss of effects enhances endothelial dysfunction, as in obese people.