Intravenously administered lorazepam (0.05 mg/kg), diazepam (0.25 mg/kg), and midazolam (0.1 mg/kg) were compared for sedation during oral surgery under local anesthesia. Sixty patients were randomly allocated into three groups in this double-blind, parallel study. The results from this trial show that all three drugs provide satisfactory sedation. Average mean arterial pressures, however, decreased significantly with midazolam and diazepam. Statistically significantly higher heart rates during the entire procedure were also found for lorazepam when compared with diazepam and midazolam. At the postblock stage, the midazolam group had respiratory rates that were significantly higher than those of the other two drug groups. Patients in the diazepam and midazolam groups took significantly longer to complete the pegboard test at the preblock stage than those in the lorzepam group. At 1, 1.5, and 2 hours after arrival in the recovery room, an inversion of groups took place, with the lorazepam group taking significantly longer for their tests than the other two groups. Significantly more improvement in anxiety levels was found at 10 minutes postdrug for the patients who had received diazepam and this tended to remain so on arrival in the recovery room. When compared with the other two groups, significantly more patients in the lorazepam group reported giddiness/dizziness and significantly more in the diazepam group reported pain on injection.