INTRODUCTION: When expressed as a percentage of the average result in young adults, bone mineral content lags behind bone length before puberty. Even though this observation has led to speculation about bone fragility in children, such relationships could simply be due to scaling effects when measures with different geometrical dimensions are compared. METHODS: The study population comprised 145 healthy subjects (6-25 years, 94 females). Magnetic resonance imaging and dual-energy X-ray absorptiometry were used to determine femur length, bone mineral content, cortical bone mineral density, cross-sectional bone geometry (bone diameter; cortical thickness; total, cortical and medullary areas; cross-sectional and polar moments of area; bone strength index) and muscle area at the proximal one-third site of the femur. Results were dimensionally scaled by raising two-, three- and four-dimensional variables to the power of 1/2, 1/3 and 1/4, respectively. Sex-differences were also assessed before and after functionally adjusting variables for femur length and weight or muscle size. RESULTS: In prepubertal children, unscaled results expressed as percentages of adult values were lowest for variables with the highest dimensions (e.g., moments of area