The effects of triazolam on cognitive function and vigilance on the morning following a nocturnal administration were investigated using event-related potentials (ERP) measurement and a sleep latency test (SLT). We previously reported a significant reducing effect on target N1 amplitude on the morning following triazolam administration, suggesting a residual effect of triazolam. In order to demonstrate, which aspect of cognitive function alteration caused the reducing effect on N1 amplitude, we added the ignore condition for ERP measurement, which enabled us to separate mismatch negativity (MMN) from other subcomponents overlapping N1. As a result, MMN was attenuated and sleep latency was shortened on the morning following triazolam administration. Two possibilities were suggested for the mechanism of MMN attenuation. One is GABAergic activation caused by the residual effect of triazolam per se, and the other is the lowered vigilance level demonstrated in the SLT. Further studies are necessary to determine whether this alteration in physiological bases underlying mismatch detection is specific to triazolam and/or other benzodiazepines or related to nonspecific vigilance level.