Abstract Background In view of accumulating case reports of thyroid dysfunction following COVID-19 vaccination, we evaluated the risks of incident thyroid dysfunction following inactivated (CoronaVac) and mRNA (BNT162b2) COVID-19 vaccines using a population-based dataset. Methods We identified people who received COVID-19 vaccination between 23 February and 30 September 2021 from a population-based electronic health database in Hong Kong, linked to vaccination records. Thyroid dysfunction encompassed anti-thyroid drug (ATD)/levothyroxine (LT4) initiation, biochemical picture of hyperthyroidism/hypothyroidism, incident Graves’ disease (GD), and thyroiditis. A self-controlled case series design was used to estimate the incidence rate ratio (IRR) of thyroid dysfunction in a 56-day post-vaccination period compared to the baseline period (non-exposure period) using conditional Poisson regression. Results A total of 2,288,239 people received at least one dose of COVID-19 vaccination (57.8% BNT162b2 recipients and 42.2% CoronaVac recipients). 94.3% of BNT162b2 recipients and 92.2% of CoronaVac recipients received the second dose. Following the first dose of COVID-19 vaccination, there was no increase in the risks of ATD initiation (BNT162b2: IRR 0.864, 95% CI 0.670–1.114; CoronaVac: IRR 0.707, 95% CI 0.549–0.912), LT4 initiation (BNT162b2: IRR 0.911, 95% CI 0.716–1.159; CoronaVac: IRR 0.778, 95% CI 0.618–0.981), biochemical picture of hyperthyroidism (BNT162b2: IRR 0.872, 95% CI 0.744–1.023; CoronaVac: IRR 0.830, 95% CI 0.713–0.967) or hypothyroidism (BNT162b2: IRR 1.002, 95% CI 0.838–1.199; CoronaVac: IRR 0.963, 95% CI 0.807–1.149), GD, and thyroiditis. Similarly, following the second dose of COVID-19 vaccination, there was no increase in the risks of ATD initiation (BNT162b2: IRR 0.972, 95% CI 0.770–1.227; CoronaVac: IRR 0.879, 95%CI 0.693–1.116), LT4 initiation (BNT162b2: IRR 1.019, 95% CI 0.833–1.246; CoronaVac: IRR 0.768, 95% CI 0.613–0.962), hyperthyroidism (BNT162b2: IRR 1.039, 95% CI 0.899–1.201; CoronaVac: IRR 0.911, 95% CI 0.786–1.055), hypothyroidism (BNT162b2: IRR 0.935, 95% CI 0.794–1.102; CoronaVac: IRR 0.945, 95% CI 0.799–1.119), GD, and thyroiditis. Age- and sex-specific subgroup and sensitivity analyses showed consistent neutral associations between thyroid dysfunction and both types of COVID-19 vaccines. Conclusions Our population-based study showed no evidence of vaccine-related increase in incident hyperthyroidism or hypothyroidism with both BNT162b2 and CoronaVac.