A growing amount of evidence suggests that thyroid-stimulating hormone (TSH) is associated with cardiometabolic risk. However, there have been few longitudinal studies. The aim of this study was to explore the causal relationship between TSH and metabolic syndrome (MetS) in a large population-based longitudinal study. From 2010 to 2016 at the Health Management Center at Tri-Service General Hospital, 25,121 eligible patients were enrolled in our cross-sectional analyses. Cox proportional hazard models were used to investigate the longitudinal association among hypertension (HTN), prediabetes (pre-DM), MetS, diabetes (DM) and TSH levels (N = 12,463). The average follow-up time was 7.2 years. In the cross-sectional analysis, the OR for MetS was 1.06 (95% CI = 1.03-1.09; P< 0.05), while the ORs for DM, pre-DM or HTN were not statistically significant (all P> 0.05). After dividing TSH levels into four quartiles, the ORs for the presence of MetS determined by comparing the highest TSH quartile with the lowest TSH quartile were 1.37 (95% CI = 1.18-1.60), 1.42 (95% CI = 1.20-1.67), and 1.44 (95% CI = 1.22-1.69) (all, P<0.05) in model 1, model 2 and model 3 respectively. The HR for the incidence of MetS was 1.33 (95% CI = 1.17-1.51; P < 0.05). Our study revealed that TSH levels had a strong association with incident MetS.