The primary purpose of this study was to examine whether alprazolam pretreatment can increase the analgesic potency of morphine without increasing opioid side-effect intensities. We employed computer-controlled, variable-rate morphine infusions based on each subject's pharmacokinetic profile for morphine derived from a tailoring bolus dose of the drug administered 1 or 2 weeks before the infusion test sessions. On each of 2 test days, we used dental electrical stimulation to determine stimulus intensity that produced consistent reports of strong (but tolerable) pain; this intensity was used for the rest of that session. Then, we measured baseline (no drug) pain intensity reports, pain-related evoked potentials recorded from vertex, and other parameters typically affected by opioids (subjective side effects). We administered alprazolam (1 mg) or placebo (lactose) orally to the subject and then repeated the test battery 30 min later. One hour after the alprazolam or placebo dose, we initiated the tailored morphine infusion to reach target plasma morphine concentration plateaus of 16, 32 and 64 ng/ml (45-min duration each) on both test days. The test battery used during baseline was then repeated at each target concentration plateau. The order of alprazolam versus placebo pretreatments was counterbalanced across subjects and known only to the investigator operating the infusion system. Results suggest that alprazolam at the dose studied did not alter analgesic potency of morphine. However, alprazolam did clearly decrease the intensity of nausea reported by subjects during and after termination of the morphine infusions. Of special interest, alprazolam alone (30 min after oral dosing) decreased evoked potential amplitude consistently without affecting pain intensity reports. This unexpected finding suggests that stimulated-related evoked potential amplitude is not an obligatory correlate of experimental pain perception.