A four degrees of freedom, damped, linear oscillator model of the lower extremity in torsion was identified through laboratory frequency response tests. The foot was rotated sinusoidally in the medial-lateral plane through the 1-20 HZ frequency band at the constant amplitudes 2, 4 and 6 degrees. Weight bearing on the foot, muscle-induced bias torsion and knee flexion were also test variables. Identification was undertaken by minimizing the squared difference between the measured and predicted transfer functions of the pelvis and foot. The model specification and the laboratory identification procedures were the focus of this work. Variations greater than 30% in the identification stiffness, damping and inertia values were predicted over the test variable ranges. The transfer functions recorded in these tests indicated the lower extremity can be modeled as a compliance between a foot inertia and the pelvis-torso inertia. Joint compression, induced by weight bearing, and muscle contraction, measured by the bias torsional moment at the foot, increased the torsional stiffness of the knee, ankle and pelvis model elements, and they had an indeterminate influence on the hip stiffness. The model knee and hip stiffnesses and the model pelvis-torso inertia were maximized, and the model ankle stiffness was minimized at the maximum foot rotation amplitude.