The effect of benzodiazepines on attention has been the object of few investigations. Studies using the spatial cueing paradigm (Posner's paradigm) have reported inconsistent results, which are likely due to methodological and/or dose differences but suggest impaired disengagement of attention from the cue to the target. The authors investigated the effect of a benzodiazepine (diazepam) on attentional shifting in the temporal domain. The attentional blink effect refers to difficulties in detecting a target if it follows the identification of a previous target occurring within a temporal window of 200-400 ms. The authors assessed whether the duration of the attentional blink was affected by diazepam. Streams of 15 real-world scenes displaying a road were presented for 50 ms each. A city name (target) appeared at Serial Positions 2, 3, or 4 of each stream. A vehicle (probe) appeared at different intervals following the city name. In a dual-task condition, participants were asked to report the city name and whether a vehicle was present. In a control condition, participants had to report only the presence of a vehicle and ignore the city name. Thirty-six healthy volunteers were assigned to 3 groups (placebo, diazepam 0.1 mg/kg, or 0.3 mg/kg). Diazepam increased both the magnitude and duration of the attentional blink effect. Participants treated with a high dose of diazepam needed more than 600 ms to detect a vehicle following identification of the name. Results suggest that diazepam at a therapeutic dosage affects attentional shifting in the temporal domain and impairs dual-task performance.