Prior studies have suggested a significant association between 25-hydroxyvitamin D (25(OH)D) concentrations with markers of inflammation and glucose and insulin homeostasis. However, it is unclear whether these associations are confounded or mediated by adiposity. We used an established mediation analysis to investigate the role of adiposity in the relation between serum 25(OH)D with markers of inflammation and glucose and insulin metabolism. We used data from National Health and Nutrition Examination Survey (2005-2010), to evaluate the associations between serum 25(OH)D and markers of insulin resistance or inflammation, and whether these associations are mediated by adiposity factors including body mass index (BMI, marker of body adiposity), waist circumference (WC, marker of central adiposity), anthropometrically predicted visceral adipose tissue (apVAT), and Visceral Adiposity Index (VAI). Analysis of co-variance and conceptual causal mediation analysis were conducted taking into consideration the survey design and sample weights. A total of 16,621 individuals were included; 8607 (48.3%) participants were men and the mean age of the population was 47.1 years. Mean 25(OH)D serum concentration for the overall population was 57.9±0.1 nmol/L with minimal differences between men and women (57.5±0.2 nmol/L and 58.2±0.2 nmol/L, respectively). After adjustment for age, sex, season and race/ethnicity, mean levels of C-reactive protein (CRP), apolipoprotein B (apo-B), fasting blood glucose (FBG), insulin, homeostatic model assessment of IR (HOMA-IR) and β cell function (HOMA-β), haemoglobin A1c (HbA1c), and 2-h glucose were lower for the top quartile of serum 25(OH)D (all p<0.001). Body mass index (BMI) was found to have significant mediation effects (to varied extent) on the associations between serum 25(OH)D and CRP, apo-B, fasting glucose, insulin, HOMA-IR, HOMA-B and HbA1c (all p<0.05). Both waist circumference and apVAT were also found to partly mediate the associations between serum 25(OH)D with CRP, FBG, HbA1c, triglycerides and HDL-cholesterol (all P < 0.05). VAI was found to have mediation effects on CRP only (p<0.001). Using a mediation model, our findings suggest that the relationship between serum 25(OH)D, insulin resistance and inflammation, may be in part mediated by adiposity. These findings support the importance of optimizing 25(OH)D status in conditions with abnormal adiposity (i.e., obesity) and treatments for the prevention of cardio-metabolic diseases affecting adipose tissue metabolism (i.e., weight loss).