Objective Several population-based studies in iodine-deficient areas have shown an association between smoking and thyroid function. There are no population-based studies about the effects of smoking in iodine-sufficient areas. We examined the effect of smoking on thyroid function and the association with iodine intake in Korea, an area with sufficient iodine intake, much more than recommended by the World Health Organization (WHO). Design Of 5018 subjects in a population-based cohort, we included 3399 who had no history of thyroid disease were not taking thyroid medication and whose blood samples were available for measurement of thyroid function. Measurements Thyroid function test, questionnaire about smoking status and dietary intake. Results Of 3399 subjects, 397(11·7%) had subclinical hypothyroidism (SCH). Female sex was an independent risk factor for SCH. Multivariate analysis in female subjects showed the following were independent risk factors for SCH: older age, positive antithyroid peroxidase (anti-TPO) antibody status and iodine intake, whereas current smoking was inversely related with SCH. However, in male subjects, only age showed a weak association with SCH. When the interaction between smoking and other risk factors was analysed, smoking showed no association with anti-TPO antibody status, whereas it showed a significant negative interaction with iodine intake (odds ratio, 0·930; 95% CI, 0·869-0·996; P = 0·037). Furthermore, the risk for SCH was observed only in the never-smoker group; however, it was abolished in current- and ex-smoker groups. Conclusion Cigarette smoking was associated with a lower prevalence of SCH in a negative interaction with iodine intake.