In the present study we investigated whether there are differences in the patellar tendon moment arm (PTMA)-knee angle relationship between pre-pubertal children and adults, and whether the PTMA length scales to relevant anthropometric measurements in the two groups. Anthropometric characteristics and the PTMA length-joint angle relationships were determined in 20 adults and 20 pre-pubertal children of both genders. The anthropometric characteristics measured were height, body mass, knee circumference, medio-lateral knee breadth, anterior-posterior knee depth, leg length, femur length and tibia length. The PTMA was quantified from magnetic resonance images using the geometric centre of the femoral condyle method, at every 5 degrees between 55 degrees and 90 degrees of knee flexion (0 degrees is full extension). Adults had a significantly greater PTMA length at all joint angles (4.2 +/- 0.4 vs. 3.6 +/- 0.3 cm at 90 degrees ; P < 0.01), with the PTMA length decreasing from knee extension to knee flexion similarly in both adults and children. There were no significant and strong correlations between the PTMA and anthropometric measures in adults for any joint angle. In contrast, the PTMA correlated and scaled with anthropometric characteristics for the children (P < 0.05, r = 0.49-0.9) at all joint angles. The PTMA length in children was most accurately predicted at 85 degrees of flexion from the equation PTMA = -0.25 + 0.083 x tibia length + 0.02 x leg length (R(2) = 0.83). These findings indicate that the knee extensor mechanism in pre-pubertal children should not be considered to be a 'scaled-down' version of that in adults.