When a fast voluntary movement is performed from a background condition of sustained antagonist muscle activation, there is often a decrease in antagonist muscle activity before the onset of the first agonist muscle burst (AG1) that continues until the onset of the antagonist muscle burst (ANT). We studied how controlling the peak velocity, movement size, and the magnitude of antagonist muscle loading affected antagonist muscle inhibition (AntI). AntI was more pronounced during movements with lower velocity and greater size, and when performed in the direction of heavier background loads, but its variation could not be related to any single kinematic or kinetic variable in all circumstances. When AntI was larger, ANT was smaller, suggesting that AntI does not play a role similar to the premotor silence of the agonist seen before AG1 when the movement is made from a background of sustained agonist contraction. When AntI was larger, the size of AG1 was also smaller, showing that, according to the motor task, different levels of reciprocal inhibition and coactivation occur at the onset of the movement. Both AG1 and AntI produce force in the direction of the desired movement, and the central nervous system selects an appropriate balance between the two, using AntI when possible.