Unexpected sensory inputs can generate a patterned startle reaction, aimed at protection and defense. Experimentally, it is usually triggered by auditory stimuli while the startle reaction to somatosensory inputs (SSS) has not received much attention so far. This may be in part due to the fact that somatosensory inputs inevitably cause local reactions, such as short and long latency reflexes and withdrawal reactions, which could interfere with recognition of the startle-related activity. Therefore, we have undertaken a study aimed at separating the SSS from other responses by exploring the responses that are common to somatosensory stimuli applied to different sites and examining the inhibitory effects of prepulse stimuli. In 13 healthy naive subjects, we applied electrical stimuli to the median nerve at the wrist (MW) or the posterior tibial nerve at the ankle (PT) and recorded from orbicularis oculi (OOC), masseter (MAS), sternocleidomastoid (SCM) and representative muscles of the limbs being stimulated (flexor carpi radialis for MW and tibialis anterior for PT). In random trials, we also applied prepulse stimuli, either a low-intensity auditory stimulus or low-intensity electrical stimuli, 100 ms before the SSS-eliciting stimulus. The pattern of SSS was different for upper and lower limb stimuli. While stimuli applied to MW induced a prominent reaction of the OOC, at a mean latency of 61.1 ms (SD = 16.3 ms), followed by the SCM at a mean latency of 83.3 ms (SD = 28.6 ms), those applied to the PT caused a small or absent response in the OOC and a consistent response of the SCM at a mean latency of 89.7 ms (SD = 30.1 ms). Prepulse stimuli effectively inhibited the responses of facial and neck muscles but only partially those of the wrist flexors to MW or the tibialis anterior to PT. Our results indicate that, although there are common neck and facial muscle reactions to somatosensory stimuli applied to upper and lower limbs, the pattern of the SSS differs according to the source of the input. Prepulse inhibition is more effective on the responses of neck and facial muscles than on the responses of limb muscles to somatosensory stimuli. These results could help in distinguishing between withdrawal and SSS reactions.