Recent evidence shows that the primary motor cortex continues to send motor commands when amputees execute phantom movements. These commands are retargeted toward the remaining stump muscles as a result of motor system reorganization. As amputation-induced reorganization in the primary motor cortex has been associated with phantom limb pain we hypothesized that the motor control of the phantom limb would differ between amputees with and without phantom limb pain. Eight above-elbow amputees with or without pain were included in the study. They were asked to produce cyclic movements with their phantom limb (hand, wrist, and elbow movements) while simultaneously reproducing the same movement with the intact limb. The time needed to complete a movement cycle and its amplitude were derived from the kinematics of the intact limb. Electromyographic (EMG) activity from different stump muscles and from the homologous muscles on the intact side was recorded. Different EMG patterns were recorded in the stump muscles depending on the movement produced, showing that different phantom movements are associated with distinct motor commands. Phantom limb pain was associated with some aspects of phantom limb motor control. The time needed to complete a full cycle of a phantom movement was systematically shorter in subjects without phantom limb pain. Also, the amount of EMG modulation recorded in a stump muscle during a phantom hand movement was positively correlated with the intensity of phantom limb pain. Since phantom hand movement-related EMG patterns in above-elbow stump muscles can be considered as a marker of motor system reorganization, this result indirectly supports the hypothesis that amputation-induced plasticity is associated with phantom limb pain severity. The discordance between the (amputated) hand motor command and the feedback from above-elbow muscles might partially explain why subjects exhibiting large EMG modulation during phantom hand movement have more phantom limb pain.