Electrical stimulation may produce excitation or inhibition of the motor neurons, as represented the blink reflex and masseter silent period in response to trigeminal nerve stimulation. Clinically, a light touch on the palm may evoke a mentalis muscle response (MMR), i.e. a palmomental reflex. In this study, we attempted to characterize the MMR to median nerve stimulation. Electrical stimulation was applied at the median nerve with recordings at the mentalis muscles. An inhibition study was done with continuous stimuli during muscle contraction (I1 and I2 of MMRaverage). Excitation was done with a single shot during muscle relaxation (MMRsingle) or by continuous stimuli during muscle contraction (E1 and E2 of MMRaverage). The characteristic differences between MMRaverage and MMRsingle were as follows: earlier onset latencies of MMRaverage (MMRaverage < 45 ms; MMRsingle > 60 ms), and a lower amplitude of MMRaverage (MMRaverage < 50 microV; MMRsingle > 150 microV). The receptive field of MMRsingle was widespread over the body surface and that of MMRaverage was limited to the trigeminal, median and index digital nerves. Series of stimuli usually significantly decreased the amplitude of MMRsingle, as a phenomenon of habituation. On the other hand, it was difficult to evoke the earlier response (i.e. MMRaverage) without continuous stimuli and an average technique. MMRaverage had the components of both excitation (E) and inhibition (I); for example, E1-I1-E2-I2 or I1-E2-I2. E2 was the most consistent component. In patients with dorsal column dysfunction, median nerve stimulation could successfully elicit MMRsingle, but not MMRaverage. Contrarily, in patients with pain sensory loss, it was more difficult to reproduce MMRsingle than MMRaverage. It seemed that MMRaverage and MMRsingle did not have equivalents across the different modalities of stimulation.