Background: Reference intervals of thyroid-stimulating hormone (TSH) and free thyroxine (FT4) are statistically defined by the 2.5^th-97.5^th percentiles, which do not account for potential risk of clinical outcomes. We aimed to define the optimal healthy ranges of TSH and FT4 based on the risk of cardiovascular disease (CVD) and mortality. Methods: We performed an individual-participant data (IPD) analysis and identified prospective cohorts with baseline TSH and FT4 measurements, CVD outcomes and mortality via the Thyroid Studies Collaboration, supplemented with a systematic literature search. The primary outcome was a composite of CVD events (heart failure, coronary heart disease, stroke) and overall mortality. CVD events, CVD mortality and overall mortality were assessed separately as secondary outcomes. We transformed TSH and FT4 into cohort-specific percentiles. We performed a one-step (cohort-stratified Cox proportional hazard models) and two-step approach (random-effects models) meta-analysis adjusting for relevant confounders. Results: We included IPD from 128,424 participants from 26 cohorts (median age 59 years, median follow-up 11.0 years). TSH in the lowest quintile was associated with a higher risk of composite outcome and overall mortality, with age- and sex-adjusted hazard ratio (HR) of 1.07 (95%CI; 1.02-1.12) and 1.09 (95%CI; 1.04-1.14) respectively, compared with the fourth quintile. There was a J-shaped association of FT4 with the composite and secondary outcomes, where FT4 in the 20^th-40^th percentiles showed the lowest risk. Compared with the second quintile, age and sex- adjusted HR for FT4 in the highest quintile was 1.20 (95%CI; 1.11-1.31) for the composite outcome, 1.32 (95%CI; 1.20-1.44) for overall mortality, 1.53 (95%CI; 1.31-1.80) for CVD mortality and 1.22 (95%CI; 1.11-1.33) for CVD events. Conclusions: TSH between 60^th and 80^th percentiles, FT4 between 20^th and 40^th percentiles represented the optimal healthy ranges defined by the risk of CVD and mortality, providing evidence for refining reference ranges of thyroid function.