For the wheelchair racing population, it is uncertain whether musculoskeletal models using the maximum isometric force-generating capacity of nonathletic, able-bodied individuals are appropriate, as few anthropometric parameters for wheelchair athletes are reported in the literature. In this study, a sensitivity analysis was performed in OpenSim, whereby the maximum isometric force-generating capacity of muscles was adjusted in 25% increments to literature-defined values between scaling factors of 0.25x and 4.0x for 2 elite athletes, at 3 speeds representative of race conditions. Convergence of the solution was used to assess the results. Artificially weakening a model presented unrealistic values, while artificially strengthening a model excessively (4.0x) demonstrated physiologically invalid muscle force values. The ideal scaling factors were 1.5x and 1.75x for each of the athletes, respectively, as was assessed through convergence of the solution. This was similar to the relative difference in limb masses between dual-energy X-Ray absorptiometry data and anthropometric data in the literature (1.49x and 1.70x), suggesting that dual-energy X-ray absorptiometry may be used to estimate the required scaling factors. The reliability of simulations for elite wheelchair racing athletes can be improved by appropriately increasing the maximum isometric force-generating capacity of muscles.