This study examined the changes in spinal and cortical evoked potentials (EPs) produced by fentanyl in the rabbit. The reversibility of these effects by naloxone and the effects of varying stimulus intensity also were examined. Eleven ketamine-anesthetized rabbits underwent analysis of spinal and cortical EPs produced by posterior tibial nerve electrical stimulation. Progressive fentanyl doses up to 100 μg/kg total were given intravenously. Stimulus intensities of two to eight times the motor threshold (MT) were used. The lowest stimulus (2 x MT) was insufficient to activate AS fibers whereas the highest (8 x MT) produced consistent AS activation. Two spinal potential waveform components (N1, P2), at 4 and 10 ms, and four cortical components (P1, N2, P3, N4), at 15, 30, 60, and 95 ms, were analyzed. Significant amplitude reductions in spinal P2 (46%), cortical N2 (54%), cortical P3 (47%), and cortical N4 (45%), P < 0.01, as well as latency prolongation in cortical P1 (11%), P< 0.01, resulted from administration of a total dose of 100 μg/kg of fentanyl. The effect of fentanyl was critically dependent on stimulus intensity for the cortical P3 component (P < 0.01). All changes were reversed by naloxone. Fentanyl produced dose-related changes in specific spinal and cortical EP waveform components in the ketamine-anesthetized rabbit. This effect was significantly dependent on stimulus intensity in at least one cortical component (P3). Such an interaction between drug dose and stimulus intensity may be relevant to interpreting other human and animal evoked potential studies.