CONTEXT: Except from associations study with body weight, there are few longitudinal data regarding the association between thyroid function and anthropometric such as waist circumference, waist-to-hip-ratio or waist-to height-ratio. OBJECTIVE: We aimed at investigating the association of thyroid stimulating hormone (TSH) at baseline with changes in different anthropometric markers between baseline and follow-up in the general population. DESIGN AND SETTING: We used data from four population-based longitudinal cohort studies and one population-based cross-sectional study. SUBJECTS: We studied 16,902 (8,204 males and 8,698 females) subjects aged 20 to 95 years from the general population. MAIN OUTCOME MEASURES: We measured body mass index, waist-circumference, waist-to-hip- ratio and waist-to-height-ratio. Multivariable median regression models were calculated adjusting for the following covariates: age, sex, baseline value of the respective anthropometric marker, smoking status, follow-up-time period and study site. RESULTS: In cross-sectional analyses, serum TSH within the reference range was positively associated with waist circumference (β 0.94 cm (95%CL 0.56; 1.32)) and waist-to-height-ratio (β 0.029 (95%CL 0.017; 0.042)). These associations were also present for the full range of TSH. In the longitudinal analyses, serum TSH at baseline was inversely associated with 5-year change of all considered anthropometric measures within the prior defined study-specific reference range, as well as in the full range of serum TSH. CONCLUSION: High TSH serum levels were positively associated with current anthropometric markers even in the study-specific reference ranges. In contrast, high TSH serum levels were associated with decreased anthropometric markers over a time span of approximately 5 years. Further research is needed to determine possible clinical implications as well as public health consequences of these findings.