BACKGROUND: The objective of this study was to investigate the relationship between smoking and metabolic syndrome in men. METHODS: This cross-sectional study included 1,852 men over age 40 who underwent health screening from April 2009 to December 2010. We classified them into three smoking levels as non-, intermediate-, and heavy-smoker, considering their smoking status (non, ex, current) and amount (0, 1-29, > or =30 pack year [PYR]). The relationship between smoking level and metabolic syndrome was analyzed by logistic regression analysis, after covariates (age, body mass index, education, house income, alcohol intake, and physical activity) were controlled. RESULTS: The proportions of non-, intermediate-, and heavy-smokers were 31.8%, 56.2%, and 12.0%, respectively. Odds ratios (ORs) and 95% confidence intervals (95% CIs) for metabolic syndrome were 1.0, 1.58 (1.09-2.23), 1.92 (1.29-2.81) in non-, intermediate-, and heavy-smokers, respectively. For heavy-smokers compared with non-smokers, ORs and 95% CIs of a lower high density lipoprotein cholesterol, higher triglyceride, and higher fasting glucose were 2.47 (1.63-3.74), 1.71 (1.17-2.52), and 1.43 (1.02-2.00), respectively. In current-smokers, we divided into three subgroups according to PYR, and compared with 1-19 PYR, ORs and 95% CIs of 20-29 PYR and > or =30 PYR for metabolic syndrome were 2.07 (1.14-3.74) and 3.06 (1.66-5.62), respectively. CONCLUSION: This study showed a positive dose-response relationship between smoking level and metabolic syndrome in men.