OBJECTIONS: Cotinine, the major metabolite of nicotine, is a useful marker of exposure to tobacco smoke and self-reporting of smoking status is thought not to be reliable. This study aimed to evaluate the agreement between the smoking self-report among adolescents and the urinary cotinine test. METHODS: The study subjects were 1226 middle and high school students in Hanam city, who were selected by stratified random sampling. The self-report about smoking behavior was compared with urine cotinine value measured with PBM AccuSignRfi Nicotine (Princeton BioMeditech Corporation, USA). The percentage agreement, kappa and 95% confidence interval (CI) were calculated. RESULTS: The overall percentage agreement was 88.6%, and those for boys, girls, middle school, general school and vocational school students were 87.3%, 90.1%, 93.7%, 85, 5%, 90.7%, and 78.4%, respectively. The overall kappa index was 0.46 (95% CI=0.39-0.54) for overall, and those for boys, girls, middle school, general school and vocational school students were 0.56 (95% CI=0.48-0.65), 0.20 (95% CI=0.07-0.32), 0.21 (95% CI=0.09-0.34), 0.55 (95% CI=0.47-0.64), 0.42 (95% CI=0.33-0.52), and 0.48 (95% CI=0.36-0.60), respectively. CONCLUSION: The percentage agreement was relatively high but the kappa values very low for girls, and middle school students. Though the prevalence bias can be influenced by these results, the selfreport was not a sufficient tool for the evaluation of adolescents' smoking status, especially in girls or middle school students.