This study assessed the somatosensory evoked potentials (SEPs) in dogs and cats to compare the effect of remifentanil on the action potentials evoked by peripheral noxious stimulation in the spinal cord. Five healthy dogs and five healthy cats underwent general anaesthesia induced with propofol and maintained with isoflurane. Each animals received all dosage of a constant-rate infusion of remifentanil at 0 (control), 0.25, 0.5, 1.0 or 2.0 μg/kg/min. The hair of the dorsal foot of a hind limb was clipped and an intraepidermal stimulation electrode that could selectively stimulate the nociceptive Aδ and C fibres was attached. An electrical stimulus was generated by a portable peripheral nerve testing device. The evoked potentials were recorded by two needle electrodes inserted subcutaneously in the dorsal midline between the lumbar vertebra: L3-L4 and L4-L5. Bimodal waveforms were obtained by electrical stimulation in control dogs and cats. The inhibitory effect of remifentanil was evaluated by comparing the changes in the N1P2 and P2N2 amplitudes. The N1P2 amplitude was depressed by remifentanil in a dose-dependent manner in dogs, but it showed no remifentanil-induced changes in cats. While the P2N2 amplitude was also depressed in a dose-dependent manner in dogs, it showed milder remifentanil-induced effects in cats. The N1P2 and P2N2 amplitudes observed herein are assumed to represent the evoked potentials derived from the Aδ and C fibres, respectively. Thus, the inhibitory effect of remifentanil on nociceptive transmission at the spinal cord was much weaker in cats, especially for transmissions possibly derived from Aδ fibres.