The aim of the present study was to clarify whether cognitive impairments caused by benzodiazepines (BDZs) are a consequence of their specific direct effects on cognitive function or whether they are explained as secondary effects of increased sleepiness. Ten healthy men (mean age, 33.9 years) participated in two experimental sessions in a randomized cross-over, double-blind study: in one session subjects were given a placebo and in the other they were given 0.125 mg triazolam (TRZ). Each experimental session was conducted on 1 day. After a pre-drug EEG recording and an event-related potential (ERP) recording, under an oddball paradigm, subjects took the TRZ or placebo orally at 1000 hours. Thereafter, EEG and ERP recording sessions, following the same procedure as the pre-drug sessions, were conducted at 1, 2, 4, 6 and 8 h after drug administration. The EEG and ERP recordings from Cz and Pz referred to the bilaterally linked ear electrodes were used. We found that P300 latency was significantly prolonged in TRZ condition at 2 h (Pz) and 4 h (Cz and Pz) after TRZ, and that the P300 amplitude was significantly reduced at 2 h (Cz and Pz) and 4 h (Pz) after TRZ, compared to the same times after placebo. The absolute power values for the theta (4-7 Hz), alpha 1 (8-9 Hz), and alpha 2 (10-12 Hz) bands did not differ at any measurement time between the treatments. Only the beta band (13-19 Hz) power value was significantly elevated after the TRZ administration (versus placebo). No significant sedative effects were detected in subjective measurements. These results indicate that a single oral dose of 0.125 mg TRZ caused cortical changes without distinct general sedation or subjective sleepiness.