Co-contraction is a common strategy when performing difficult or unstable motor tasks. For example, we co-contract muscles of the arm when generating downward force to turn a screw and co-contract muscles of the leg when walking on a slippery surface. It is thought that the primary benefit of co-contraction is to directly increase the mechanical impedance (stiffness and damping) of muscle to generate greater instantaneous restoring forces when a limb or joint is perturbed. However, this explanation ignores the role of neural feedback in corrective responses. Characterizing the effects of co-contraction on motor function is important to understanding its use in common tasks and its role in pathologies such as knee osteoarthritis. A series of experiments were performed to investigate the effects of co-contraction on the motor response to physical perturbations. First, participants performed an upper limb postural control task in which a random torque pulse perturbed the elbow. Muscle activity was increased in the elbow flexors or extensors independently by resisting a background load or in both muscle groups simultaneously by co-contracting. Rapid corrective responses were generated without overshoot even at the lowest level of cocontraction, whereas activating a single muscle group more than double that amount resulted in poorer performance. Thus, it was not the total muscle activity, and thus the increase in limb impedance, that improved performance when co-contracting, but the fact that both muscle groups were simultaneously pre-activated before the disturbance. It was shown that the primary benefit of co-contraction was the ability to engage both the stretched and shortened muscles in the corrective response. This behavior was also observed in an upper limb tracking task, demonstrating that it was not a simple strategy invoked only for stationary postural control, but a strategy that also improved performance during movement. Further work expanded the generalizability of this behavior in the upper limb and extended it to the lower limb. Finally, the benefits of this strategy were examined using an optimal feedback control model. The results of this work indicate that co-contraction engages a unique motor strategy, providing new insight into its role in motor function.