There is no practical or ethical way to measure the forces, stresses, and strains that occur inside the human body during activity. Musculoskeletal modeling enables estimation and prediction of these quantities. This information allows the investigation of treatments designed to address orthopedic pathology and prevent injury. Musculoskeletal modeling is a predictive computational approach that consists of detailed representations of the bones, muscles and ligament anatomy, driven in simulation by measurements of subject specific mechanics. During normal physical activity, the muscles of the human body generate the majority of the load applied to the joints and passive soft tissues such as the ligaments. Muscle forces are predicted through representation of the geometry and force-generating properties of the muscles, and combination with optimization to predict muscle coordination strategies during movement. Our current work focuses on the development and demonstration of a multi-scale musculoskeletal model of the human body in a single framework that is capable of realistic simulation of dynamic physical activity. This framework is applied to current issues in total knee replacement, hip fracture, and spinal stability. The overall vision of this initiative is to fill a broad need in medical research for more realistic representation of the human tissues in simulation.