We examined the age- and sex-specific distributions of biomarkers of tobacco smoke exposure to determine the optimal cutoffs to distinguish smokers from non-smokers over the last 10 years in Korea, during which smoking prevalence and secondhand smoke (SHS) exposure declined due to changes in tobacco control policy. We analyzed data from the Korea National Health and Nutrition Examination Survey on creatinine-adjusted urinary cotinine (2008-2018; 33 429 adults: 15 653 males and 17 776 females) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL; 2016-2018; 6337 adults: 3091 males and 3246 females). We determined the optimal cutoffs and confidence intervals (CIs) to distinguish smokers from non-smokers using receiver operator characteristic curve analysis and bootstrapping (1000 resamples). The optimal cutoff values of creatinine-adjusted urine cotinine and NNAL concentration were 20.9 ng/mg (95% CI: 20.8-21.0, sensitivity: 96.6%, specificity: 93.8%) and 8.9 pg/mg (95% CI: 8.8-8.9, sensitivity: 94.0%, specificity: 94.7%), respectively, in 2016-2018. The optimal cutoffs of both biomarkers increased with age and were higher in females than in males for NNAL concentration. In both sexes, the optimal cutoff of urine cotinine continuously declined over the study period. The optimal cotinine cutoff declined along with smoking prevalence and levels of SHS exposure due to enforcement of tobacco control policies, including smoke-free ordinances and tax increases. Monitoring of biomarkers of tobacco exposure appears necessary for verification of smoking status and regulatory use. Our results based on nationally representative data suggest that a large decrease in the optimal cutoff value of urine cotinine to distinguish smokers from non-smokers was caused by decreases in smoking prevalence and SHS exposure following enforcement of tobacco control policies over the last 10 years. We determined the optimal cutoff values of urine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), which were not previously reported in representative population in Asia, to enable more accurate estimation of exposure to tobacco smoke and proper assessment of disease risks. Gender- and age-specific differences in the optimal cutoffs require further study. Monitoring of biomarkers of tobacco smoke exposure seems necessary for verification of smoking status and regulatory use.