High time resolution is required to reliably measure neuronal activity in the gustatory cortex in response to taste stimuli. Hedonic aspects of gustatory processing have never been explored using gustatory evoked potentials (GEPs), a high-time-resolution technique. Our aim was to study cerebral processing of hedonic taste in humans using GEPs in response to sucrose solutions in subjects with different ratings of pleasantness regarding sucrose. In this exploratory study, 30 healthy volunteers were randomly stimulated with 3 sucrose solutions. The sucrose stimulus was presented to the tongue for 1s 20 times. GEPs were recorded from 9 cortical sites with EEG sensors at Cz, Fz, Pz, C3, C4, F3, F4, Fp1, and Fp2 (10/20 system). The main result was that subjects who preferred the high-concentration (20g/100mL) sucrose solution had higher GEP amplitudes on the Pz, Cz, and Fz electrodes than did subjects who preferred the low-concentration (5g/100mL) or the moderate-concentration (10g/100mL) solutions regardless of stimulus intensity. The difference in P1N1 amplitude on the Pz, Cz, and Fz electrodes according to sucrose preference of the subjects was described with stronger significance with stimulation by the 20 g-sucrose solution than by the 5 and 10g sucrose solutions. Using the reliable and safe GEP technique, we provide an original demonstration of variability of the gustatory response on the Pz, Cz, and Fz electrodes according to a sweet preference in humans. Further studies are needed to correlate the electric signal recorded by surface electrodes to the neural generator.