A common method designed to identify if an animal hears tinnitus assumes that tinnitus “fills-in” silent gaps in background sound. This phenomenon has not been reliably demonstrated in humans. One test of the gap-filling hypothesis would be to determine if gap-evoked cortical potentials are absent or attenuated when measured within background sound matched to the tinnitus sensation. However the tinnitus sensation is usually of low intensity and of high frequency, and it is unknown if cortical responses can be measured with such “weak” stimulus properties. Therefore the aim of the present study was to test the plausibility of observing these responses in the EEG in humans without tinnitus. Twelve non-tinnitus participants heard narrowband noises centered at sound frequencies of 5 or 10 kHz at sensation levels of either 5, 15, or 30 dB. Silent gaps of 20 ms duration were randomly inserted into noise stimuli, and cortical potentials evoked by these gaps were measured by 64-channel EEG. Gap-evoked cortical responses were statistically identifiable in all conditions for all but one participant. Responses were not significantly different between noise frequencies or levels. Results suggest that cortical responses can be measured when evoked by gaps in sounds that mirror acoustic properties of tinnitus. This design can validate the animal model and be used as a tinnitus diagnosis test in humans.