Background: The number of CAG repeats (CAG n ) within the CAG repeat polymorphism of the androgen receptor gene correlates inversely with the transactivation of the receptor. Objective: To examine the impact of CAG n on muscle, fat distribution, and circulating androgen levels. Design, settings and participants: Population-based, cross-sectional study of 783 Danish men aged 20-29 years. Methods: Genotyping was performed in 767 men. Areas of thigh and lower trunk muscle (muscle thigh and muscle lower trunk ), subcutaneous adipose tissues (SAT thigh and SAT lower trunk ), and deep adipose tissues (i.m. and visceral) were measured in 393 men by magnetic resonance imaging (MRI). Lean body mass (LBM) and fat mass (FM) were measured in all men by whole body dual-energy X-ray absorptiometry (DEXA). The absolute areas acquired by MRI were the main outcomes. The absolute DEXA measurements and relative assessments of both modalities were considered as the secondary outcomes. Results: CAG n (range: 10-32) correlated inversely with absolute muscle thigh (rZK0.108), absolute muscle lower trunk (rZK0.132), relative muscle thigh (rZK0.128), relative muscle lower trunk (rZK0.126), relative LBM lower extremity (rZK0.108), and relative LBM total (rZK0.082), and positively with relative SAT thigh (rZ0.137), relative SAT lower trunk (rZ0.188), relative FM lower extremity (rZ0.107), and relative FM total (rZ0.082). These relationships remained significant, controlling for physical activity, smoking, chronic disease, and age. CAG n did not correlate with any circulating androgen. Conclusions: The CAG repeat polymorphism affects body composition in young men: absolute muscle thigh and absolute muscle lower trunk increase as CAG n decreases. Expressed relatively, muscle areas and LBM increase, while SAT and FM decrease as CAG n decreases. The polymorphism does not affect deep adipose tissues or circulating androgen levels in young men.