Dynamic joint stiffness defines the dynamic relationship between the position of the joint and the torque acting about it; hence it is important in the control of movement and posture. Joint stiffness consists of two components: intrinsic stiffness and reflex stiffness. Measuring intrinsic and reflex torques directly is not possible, thus estimating intrinsic and reflex stiffness is challenging. A further complication is that both intrinsic and reflex stiffness vary with joint position and torque. Thus, the measurement of dynamic joint stiffness during movement requires a time-varying algorithm. Recently we described an algorithm to estimate time-varying intrinsic and reflex stiffness and demonstrated its application. This paper describes modifications to that algorithm that significantly improves the accuracy of the estimates it generates while increasing its computational efficiency by a factor of seven.