Rather than relying solely on subjective pain evaluation using means such as the visual analogue scale (VAS), in clinical situations it is possible to observe evoked responses of the autonomic nervous system (ANS) as objective indicators. Few studies, however, have reported these relationships under finely controlled sedation. 16 healthy male participants were administrated in intravenous sedation with either propofol or midazolam randomly. We initially determined, using pharmacokinetic simulation, the effect-site concentration (Ce) of anaesthetic at loss of response to verbal command and eyelash reflex (Ce-LOR). Then subsequently adjusted Ce to 75%, 50%, and 25% of Ce-LOR to achieve deep, moderate, and light sedation. At awake control state and each sedation level, a noxious electrical stimulation was applied three times at the right forearm, an average pain intensity of the three stimuli was rated on a VAS (0-10). Changes in the peripheral perfusion index measured by oximetry were used as an indicator of ANS response. We analyzed the influence of sedation level on VAS and ANS responses compared to the awake control state. While ANS responses were similar in all conditions, VAS was statistically significantly lower in moderate (5.6±0.6, p <0.005) or deep (5.3±0.6, p <0.001) sedation than in the awake state (7.2±0.4). This study revealed that even when the ANS responds similarly to the same stimulation, subjective pain perception is attenuated by sedation. A cerebral mechanism other than that of the brainstem might determine subjective pain intensity.