This study assessed whether experimentally determined narcotic analgesia in human subjects represents a pharmacologic effect or a psychological effect of detecting the administration of an active medication. Forty dental patients used a verbal descriptor procedure to assess both the intensity and unpleasantness of sensations produced by electrical stimulation of intact teeth. Stimuli were rated before and after an intravenous injection of 0.11 mg/kg of diazepam, to produce detectable side effects in all patients, followed by a double-blind intravenous injection of either 0.66 μg/kg of fentanyl or saline placebo. The results were similar to previous findings in which diazepam was not administered: only intensity responses were reduced after fentanyl administration and only unpleasantness responses were reduced after placebo administration. These results suggest that the reduction in pain intensity following fentanyl administration represents an analgesic effect and not an artifact of detecting the administration of an active medication. They also suggest that diazepam at this dose does not alter pain sensations produced by electrical tooth pulp stimulation.