Despite the benefits of exercise for bone health, the development of protocols to improve bone quality in fracture prone areas remains a challenge because of the difficulty in determining how bone is loaded in vivo. Discussions persist regarding whether the femoral neck is in bending or compression and the degree to which muscle forces contribute to loading. We simulated the loads during walking using subject-specific muscle-driven finite element models and compared three loading configurations:​ [1] all hip-spanning muscle and joint forces, [2] a sub-set of hip muscle and joint forces, and [3] hip joint forces only. We found that the use of the hip joint contact force alone resulted in large tensile and compressive principal strains. However, including muscle forces resulted in significantly decreased strain magnitudes and a change of mode to uneven compression in the femoral neck suggesting that the inclusion of muscle forces offset the bending moment observed in a hip-joint-force-only model. In the pertrochanter, including hip muscle forces increased the compressive and tensile strains. Together, these findings indicate that including hip-spanning muscle forces changes the mode and distributions of strains in the femoral neck as well as strain magnitudes in the pertrochanter. Future studies will benefit from including both muscle and joint reaction forces in simulations of bone under in vivo loading conditions in order to develop informed exercise interventions to improve bone health.