Humans can compensate rapidly to unforeseen errors and circumstances while performing motor tasks. These tasks are exposed to sensory and motor noise and delays originated from the biological system characteristics, as well as to varying surrounding environments while performing movements. Therefore, uncertainty of the estimate of limb position arises. Error estimates during movement and displacements after a mechanical perturbation are compensated by relying on proprioceptive feedback from the body. Moreover, the motor system can rely on co-contraction as it increases the amplitude of the short latency stretch reflex, which plays an important role in minimizing the effects of disturbances. This study examines the influence of position, velocity, and pre-perturbation background muscle activity in the decision process of avoiding obstacles in the environment after a mechanical perturbation while performing a reaching task. After the perturbation, participants had to choose between two strategies: going in between or around the obstacles to reach an end target. Position of the hand, velocity, and EMG activity of four muscles in the shoulder and elbow were compared at different time epochs between both strategies. No significant differences were found in muscle activity pre-perturbation and lateral position before and up to 100ms after perturbation onset. Significant difference in lateral velocity was found 50ms after perturbation onset between the two strategies. Online corrections to avoid obstacles after a mechanical perturbation are modulated by the lateral velocity of the limb.