The regulation by muscle spindles of jaw-closing muscle activity during mastication was evaluated in anesthetized rabbits. Simultaneous records were made of the discharges of muscle spindle units in the mesencephalic trigeminal nucleus, masseter and digastric muscle activity (electromyogram [EMG]), and jaw-movement parameters during cortically induced rhythmic jaw movements. One of three test strips of polyurethane foam, each of a different hardness, was inserted between the opposing molars during the jaw movements. The induced rhythmic jaw movements were crescent shaped and were divided into three phases: jaw-opening, jaw-closing, and power. The firing rate of muscle spindle units during each phase increased after strip application, with a tendency for the spindle discharge to be continuous throughout the entire chewing cycle. However, although the firing rate did not change during the jaw-opening and jaw-closing phases when the strip hardness was altered, the firing rate during the power phase increased in a hardness-dependent manner. In addition, the integrated EMG activity, the duration of the masseteric bursts, and the minimum gape increased with strip hardness. Spindle discharge during the power phase correlated with jaw-closing muscle activity, implying that the change in jaw-closing muscle activity associated with strip hardness was caused by increased spindle discharge produced through insertion of a test strip. The increased firing rate during the other two phases may be involved in a long-latency spindle feedback. This could contribute to matching the spatiotemporal pattern of the central pattern generator to that of the moving jaw.