A measurement system was employed to detect gustatory evoked potentials from human scalp by stimulus of a taste solution with the use of a laser beam device. The evoked potentials for four taste qualities (i.e., sweet-sucrose, salty-sodium chloride, sour-tartaric acid, and bitter-quinine-HCl) were measured before and after treatment with a sweet suppressing agent (i.e., gymnema sylvestre extract) to the tongue of a human. The solution was given to the chorda tympani nerve located 20 mm from the apex of the tongue and 15 mm from the left side of the center line. The maximum potential level and its latency were evaluated. Artificial saliva was used as a control solution. The evoked potentials obtained were averaged by eight evoked potentials to detect the peak of the evoked potential more clearly. The latencies for taste stimuli were found on two kinds of peaks at approximately 50 ms and 180 ms. These peaks are P1 and P2. The purpose of this study is to investigate the influence of sweet suppressing agent on P1 and P2. The influence of the sweet suppressing agent to evoked potential by salty, sour, and bitter taste stimuli was not recognized, but the responses to sweet (sucrose) were abolished after treatment with a sweet suppressing agent. It was recognized that the peak P2 originated from the taste stimulus. The peak P1 did not suffer the influence of the sweet suppression, so it was considered that the response to P1 was due to sensations other than the gustatory response, such as somatosense.