Leg length and relative leg length are considered to be reliable markers of prepubertal living conditions. Cessation of leg growth, driven by estrogen, occurs earlier in puberty in girls than boys. We hypothesized that leg length and relative leg length, as sitting height to leg ratio, might have sex-specific associations with age of puberty. We used multivariable linear regression in 10,046 older (>or=50 years) Chinese from the Guangzhou Biobank Cohort Study (Phase 3) to examine the associations of recalled age of puberty (women: age of menarche, and men: mean age of first nocturnal emission, voice breaking, and first pubic hair) with subischeal leg length, sitting height to leg ratio, and sitting height. Leg length and sitting height to leg ratio had different associations with age of puberty in men and women (P-values for interaction <0.001), but sitting height did not. Per year earlier puberty, legs were longer among men by 0.09 cm (95% confidence interval (CI) 0.01-0.18) and shorter among women by -0.16 cm (95% CI -0.20 to -0.12). Further adjustment for age, hip size (as a marker of buttock fat), and several markers of childhood conditions did not obviate the difference in association by sex. Adult leg length and relative leg length (sitting height to leg ratio) may be biomarkers of different exposures in men and women, with corresponding implications for their interpretation as a biomarker of early life exposures. (c) 2010Wiley-Liss, Inc.