Background: Body mass index (BMI) inadequately captures body composition profiles that influence cardiometabolic disease risk. We investigated sex-specific associations of body composition with plasma lipids and lipoproteins.Methods: Blood plasma collected from the Busselton Healthy Ageing Study (n = 1,886, 53.9% female, aged 45–67 years) underwent comprehensive lipidomic and lipoprotein phenotyping, alongside analysis of supramolecular phospholipid composite (SPC) and glycoprotein (Glyc) signals. Sex-stratified partial Spearman's rank correlations, adjusted for age, blood pressure, BMI, and lipid-lowering medication use, were used to find lipidomic associations with dual X-ray absorptiometry (DXA)-derived measures of fat and lean mass.Findings: BMI adjustment inverted fat-lean correlations (fat–lean: males ρ = −0.87; females ρ = −0.89), revealing relationships masked by anthropometric measures. Visceral adipose tissue (VAT) showed the strongest lipidomic associations in both sexes: positive correlations with diacylglycerols, triacylglycerols (ρ = 0.28–0.29, p < 2.24×10⁻¹⁷), VLDL, IDL, small LDL, HDL-4, SPC and Glyc; inverse associations with lactosylceramides and large HDL. Leg lean mass exhibited opposing, albeit weaker, associations with VAT in both sexes. Notably, female leg fat correlated with a distinct favourable profile: inverse associations with glycerolipids and small LDL, and positive associations with sphingomyelins and lactosylceramides.Interpretation: DXA-derived body composition measures associate with sex-specific lipid signatures. VAT drives adverse lipidomic profiles in both sexes, whilst leg lean mass and, distinctly in females, leg fat mass associate with metabolically favourable signatures.