ABSTRACT: INTRODUCTION:Associations between smoking, cancer and mortality are well established. Although cancer mortality rates have decreased in recent years, the economic burden of smoking-related cancers continues to increase. This study investigates the economic costs of cancers related to smoking in Korea in 2014. METHODS:Cancer patients were identified through National Health Insurance Services medical claims with ICD-10 cancer codes. We multiplied the costs by the population attributable fraction for each type of cancer and calculated direct and indirect costs, where direct costs comprise direct medical and non-medical costs of inpatients and outpatients, and indirect costs include estimates of future income loss due to premature death, productivity loss during hospitalization and outpatient visits, and job loss. RESULTS:In 2014, there were 79297 smoking-related cancer patients, accounting for 8.47% of all Korean cancer cases for that year. The direct cost of cancers due to smoking was approximately 595 million USD, whereas indirect costs were much higher, at nearly 2.2 billion USD. The average expenditure of a typical patient was 34815 USD. Lung, liver and stomach cancers were most prevalent and represented the most significant share of the economic burden, whereas the largest per-patient spending was for pancreatic, liver, and lung cancers. Lung, liver and stomach cancers had the highest economic impact on men, while lung, liver and ovarian cancers had the most significant economic impact on women. CONCLUSIONS:It is imperative that more stringent steps be taken to reduce the huge economic burden of cancers linked to smoking. ABBREVIATIONS:IARC: International Agency for Research on Cancer, PAF: population attributable fraction, NHIS: National Health Insurance Services, ICD-10: International Statistical Classification of Diseases and Related Health Problem 10th Revision, GDP: Gross Domestic Product, DALYs: Disability-Adjusted-Life-Years, WHO: World Health Organization, FCTC: Framework Convention on Tobacco Control.