Temporomandibular (TM) disorders have an uncertain aetiology. The purpose of this study was to replicate and extend the work of Yemm (1969) (Archs oral Biol. 14, 873-878, 1437-1439; Br. dent. J. 127, 508-510), who reported a lack of habituation to experimentally induced stress in subjects with TM disorders. In this study, 20 TM disorder patients and 20 controls, matched for age and sex, performed a timed psychomotor task while bilateral masseter and anterior temporalis electromyographic (EMG) activity, finger temperature and skin admittance were monitored. These data were sampled, stored every 4s and averaged by an Apple II Plus/ISAAC interface. A significant trials effect for finger temperature (F[4/152] = 34.99, p less than 0.001) and skin admittance (F[4/152] = 41.90, p less than 0.001) was found, suggesting that the independent variable (stress) had been successfully manipulated. A significant trials by groups effect was found for right temporalis EMG activity (F[3/144] = 3.94, p less than 0.05); the left temporalis showed a similar, but not significant, trend. The masseter muscles did not show differences or trends between groups. No significant differences were found in resting EMG levels or in the initial magnitude of EMG responses during baseline. These results provide support for the hypothesis that TM disorder and control groups respond differently to stress in terms of habituation to stressful stimuli.