Evidence associating serum 25-hydroxyvitamin D (25(OH)D) concentrations and cardiometabolic risk factors is inconsistent and studies have largely been conducted in adult populations. We examined the prospective associations between serum 25(OH)D concentrations and cardiometabolic risk factors from adolescence to young adulthood in the West Australian Pregnancy Cohort (Raine) Study. Serum 25(OH)D concentrations, BMI, homoeostasis model assessment for insulin resistance (HOMA-IR), TAG, HDL-cholesterol and systolic blood pressure (SBP) were measured at the 17-year (n 1015) and 20-year (n 1117) follow-ups. Hierarchical linear mixed models with maximum likelihood estimation were used to investigate associations between serum 25(OH)D concentrations and cardiometabolic risk factors, accounting for potential confounders. In males and females, respectively, mean serum 25(OH)D concentrations were 73·6 (sd 28·2) and 75·4 (sd 25·9) nmol/l at 17 years and 70·0 (sd 24·2) and 74·3 (sd 26·2) nmol/l at 20 years. Deseasonalised serum 25(OH)D3 concentrations were inversely associated with BMI (coefficient −0·01; 95 % CI −0·03, −0·003; P=0·014). No change over time was detected in the association for males; for females, the inverse association was stronger at 20 years compared with 17 years. Serum 25(OH)D concentrations were inversely associated with log-HOMA-IR (coefficient −0·002; 95 % CI −0·003, −0·001; P<0·001) and positively associated with log-TAG in females (coefficient 0·002; 95 % CI 0·0008, 0·004; P=0·003). These associations did not vary over time. There were no significant associations between serum 25(OH)D concentrations and HDL-cholesterol or SBP. Clinical trials in those with insufficient vitamin D status may be warranted to determine any beneficial effect of vitamin D supplementation on insulin resistance, while monitoring for any deleterious effect on TAG.