Abstract. Objective: To measure the association between smoking and serum adiponectin, taking into consideration insulin resistance and obesity. Material and Methods: The cross-sectional study was carried out in Seoul, Korea in 2006. Waist circumference (WC), body mass index (BMI), and serum adiponectin were measured in 2,500 healthy Korean men. Multiple linear regression models were used to assess the association of smoking status with serum adiponectin level. WC, BMI, and homeostasis model assessment (HOMA) were classified into two groups according to median values. Results: The mean adiponectin concentrations were 6.6 μg/ml and 7.3 μg/ml in current smokers and non-smokers. After adjusting for age, BMI, and alcohol consumption, mean log adiponectin levels decreased by 0.064 μg/ml in current smokers compared with non-smokers (P = 0.0190). Mean log adiponectin levels also decreased by 0.030 and 0.095 μg/ml in moderate and heavy smokers compared to non-smokers. The relationship between adiponectin and smoking was similar between the high and low insulin resistance, BMI, and WC groups. Conclusions: These results suggest that serum adiponectin levels are associated with smoking status. These data also support that lower serum adiponectin concentrations in smokers may not be dependent on insulin resistance status or obesity. [7] showed that plasma adiponectin concentrations were reduced among individuals with obesity, diabetes mellitus, or coronary heart disease [7,8], implying that adiponectin plays a role in these conditions.Many studies have reported on hypoadiponectinemia in smokers [9][10][11][12][13]. However, the underlying mechanisms of this relationship are not yet fully understood. Because the prevalence of insulin resistance may also be increased in smokers [14,15], it is not clear whether hypoadiponectinemia in these individuals is due to smoking, or to the coexistence of insulin resistance. However, relevant literature on the relationship between serum adiponectin and smoking status independent of insulin resistance is sparse. Thus, the purpose of this study was to measure the association between smoking and adiponectin, taking into consideration the effects of insulin resistance and obesity. We hypothesized that adiponectin in smokers is associated with smoking status regardless of insulin resistance and obesity.