In a cohort of 190,599 participants from The National Health Insurance Service-National Health Screening (NHIS-HEALS) study, we investigated the association of changes in the predicted body composition and metabolic profiles with the risk of metabolic syndrome (MetS) in the general population, which was hitherto incompletely elucidated. At baseline and follow-up examinations, the body composition, including lean body mass (LBM), body fat mass (BFM), and appendicular skeletal mass (ASM), were estimated using a prediction equation, and the risk of MetS was analyzed according to relative body composition changes. An increase in relative LBM and ASM decreased the risk of MetS in men and women (adjusted odds ratio (aOR), 0.78 and 0.80; 95% confidence interval (CI), 0.77–0.79 and 0.79–0.81, respectively; all p < 0.001). As relative LBM and ASM increased, the risk of MetS was more significantly reduced in the group with higher baseline BMI and body fat mass index (BFMI)(all p-trend < 0.001). In men, when the relative LBM increased (aOR, 0.68; 95% CI, 0.63–0.73), the risk of MetS was low despite increased BMI. Thus, our findings suggested that an increase in the relative LBM and ASM reduced the risk of MetS, whereas an increase in the relative BFMI increased the risk of MetS; this result was consistent in men despite an increase in BMI.