In this article, a 3-link kinematic model of a human leg is defined and analyzed with focus on optimizing the manipulability. The forward kinematics for the leg is used to define quantitative measures of the manipulability and workspace in a sagittal plane. Analytical results for different manipulability indices are derived. Using numerical optimization in Matlab package, the manipulability measure is optimized under different constraints. The range of motion and joint comfort zones of every joint is defined. The algorithm for redundant chain, based on analytical equations, is proposed in inverse kinematics.