In a previous work, we studied the mechanical and the metabolic profiles of fatigue of paralyzed quadriceps muscle under activation by FES. The metabolic state of the muscle during stimulation of paraplegic patients was monitored, simultaneously with the decaying force, by using 31P nuclear magnetic resonance spectroscopy. In the present work, a musculotendon model was developed to enable prediction of the force output during continuous electrical stimulation. The model consisted of five elements, including the tendon, the parallel elastic, contractile and damper muscle elements, as well as the muscle mass. The mechanism of the contractile element was based upon the length-tension and the velocity-tension curves, the activation trajectory, and the experimentally obtained relationship between force and intracellular pH. In the equations obtained, three sets of parameters were used: 1) general muscle parameters, associated with the length-tension curves of tendon, fascia, and muscle and the velocity-tension curve of the contractile element; 2) specific anthropometric parameters of the muscle; and 3) fatigue parameters which were obtained from our previously recorded experimental data. The model was formulated to allow prediction of the quadriceps muscle force under dynamic activation and at various levels of stimulation. The model solution was for isometric contraction in supermaximal stimulation, and it provided the force decaying profiles, which were compared to those obtained experimentally. The parameters yielding the best fit between the model and the experimental results were indicated. Particularly, two muscle nonspecific parameters, namely, the muscle stress parameter and the parameter representing the ratio between the muscle's slack length and its length in vivo at various knee angles, were determined using the model. The muscle stress parameter was found to be between 60 and 64 N/cm2, and the length ratio was 0.952, 0.935, 0.920, and 0.901 for the 0, 30, 60, and 90 degrees knee angle, respectively. Finally, a sensitivity analysis was conducted of the model to perturbations of these two estimated parameters, revealing that the model was sensitive to these parameters.