The purpose of this study was to understand how stretch-related sensory feedback from an antagonist muscle affects agonist muscle output at different contraction levels in healthy adults. Ten young (25.3 ± 2.4 years), healthy subjects performed constant isometric knee flexion contractions (agonist) at 6 torque levels: 5%, 10%, 15%, 20%, 30%, and 40% of their maximal voluntary contraction. For half of the trials, subjects received patellar tendon taps (antagonist sensory feedback) during the contraction. We compared error in targeted knee flexion torque and hamstring muscle activity, with and without patellar tendon tapping, across the 6 torque levels. At lower torque levels (5%, 10%, and 15%), subjects produced greater knee torque error following tendon tapping compared with the same torque levels without tendon tapping. In contrast, we did not find any difference in torque output at higher target levels (20%, 30%, and 40%) between trials with and without tendon tapping. We also observed a load-dependent increase in the magnitude of agonist muscle activity after tendon taps, with no associated load-dependent increase in agonist and antagonist co-activation, or reflex inhibition from the antagonist tapping. The findings suggest that at relatively low muscle activity there is a deficiency in the ability to correct motor output after sensory disturbances, and cortical centers (versus sub-cortical) are likely involved.