The impact of serum thyroid hormone levels on thyroid cancer risk is unclear. Some studies reported that elevated thyroid-stimulating hormone (TSH) is associated with higher risk for incidence of thyroid cancer, but other studies reported no relationship. We conducted a large cohort study in 164,596 South Korean men and women who were free of thyroid cancer at baseline and underwent health examination with hormone levels of thyroid function. A parametric proportional hazard model was used to evaluate the adjusted hazard ratio (HR) and 95% CI. During 2,277,749.78 person-years of follow-up, 1280 incident thyroid cancers were identified (men = 593, women = 687). Among men, the multivariable-adjusted HR (95% CI) for thyroid cancer comparing low levels of TSH with normal levels of TSH was 2.95 (1.67–5.23), whereas the corresponding HR (95% CI) in women was 1.5 (0.88–2.55). High levels of free T4 and free T3 were also associated with incident thyroid cancer in both men and women. In clinical implication, overt hyperthyroidism is associated with thyroid cancer in both men and women. Within the euthyroid range, the highest tertile of TSH was associated with a lower risk of thyroid cancer than the lowest TSH tertile and the highest FT4 tertile was associated with a higher risk of thyroid cancer than the lowest FT4 tertile in both men and women. Our finding indicates that low levels of TSH and high levels of FT4, even within the normal range, were associated with an increased risk of incident thyroid cancer.