Context: In vitro and in vivo experiments have assigned both oncosuppressive and oncogenic properties to thyroid hormones. Population-based studies have found inconclusive results. Objective: We aimed to prospectively assess the relation between thyroid function and incident cancer in a population-based setting. Design, Setting, and Participants: The current study is a prospective population-based cohort study including 10 318 participants for whom baseline measurements of free T4 (FT4) and/or TSH were available. Main Outcome Measures: Cox proportional hazards models were used to assess hazard ratios (HRs) of any solid non-skin cancer, as well as lung, breast, prostate, and gastrointestinal cancer specifically. Results: Higher FT4 levels were associated with a higher risk of any solid cancer (HR, 1.42; 95% confidence interval [CI], 1.12–1.79), lung cancer (HR, 2.33; 95% CI, 1.39–3.92) and breast (HR, 1.77; 95% CI, 1.10–2.84) cancer. The risk estimates were similar after exclusion of thyroid-altering medication, but the association lost significance for breast cancer. Compared with the lowest FT4 tertile, the highest tertile was associated with a 1.13-fold increased risk of any solid, 1.79-fold increased risk of lung, and 1.14-fold increased risk of breast cancer (P for trend <.05 for all). For TSH levels we found no associations with cancer risk. There was no differential effect of sex or age on the association between thyroid function and cancer risk. Conclusions: Higher FT4 levels are significantly associated with an increased risk of any solid, lung, and breast cancer. Further research should elucidate the underlying pathophysiological mechanisms.