It has recently been shown that a slow stretch evokes a short-latency (probably monosynaptic) and a long-latency (polysynaptic) reflex response in human jaw-closing muscles. The effect of nociceptive muscle input on the fusimotor system has not been investigated in detail. In order to investigate the effect of sustained muscle pain on the jaw stretch reflex, two main experiments were performed. Stretch reflex responses were evoked in the masseter and temporalis muscles by slow stretches (1-mm displacement, 40-ms ramp time) before, during and 15 min after a period of experimentally induced muscle pain. In experiment I, a dose of 1.0 M hypertonic or 154 mM isotonic (control) saline was infused in random order into the left masseter for up to 15 min (n=12). The level of excitation of the left masseter at 15% maximal voluntary contraction was controlled by visual feedback of the surface EMG (sEMG). In experiment II, a dose of 1.0 M saline was infused into the left masseter but with feedback from the sEMG of the right masseter (n=12). In a control experiment, both sEMG and intramuscular EMG (imEMG) were recorded from the left and right masseters; the feedback was from imEMG of the left masseter (n=12). The early (onset: 9--10 ms) and late (duration from 25 to 40 ms) reflex components were recorded and analysed in all experiments. Infusion of 1.0 M saline caused moderate pain (mean score on a Visual Analogue Pain Scale: 4.9--5.0 cm). The peak-to-peak amplitude of the early reflex component in the painful masseter normalized to the pre-stimulus EMG activity was significantly higher during the pain than the pre- and post-infusion conditions in all experiments. The normalized area of the late reflex component in the painful masseter was significantly larger than in the pre-infusion condition in all experiments. Isotonic saline had no significant effect on the jaw stretch reflexes. These results indicate that experimental jaw-muscle pain in humans facilitates the early as well as the late component of the jaw stretch reflex response as revealed by both sEMG and imEMG. This effect appears to be independent of the level of excitation of the muscle and not related to volume effects of the injected saline. A change in the sensitivity of the fusimotor system during muscle pain is suggested as an explanation.