The somatosensory-evoked potential (SEP) elicited by high-intensity stimulation potentially provides a reliable indicator of analgesic efficacy since it reflects the level of activation of the nociceptive system. In the present study, components in the 10-30-ms latency range of SEPs recorded over the primary somatosensory cortex (SI-SEPs) and vertex (Vx-SEP) in the rat were characterized and compared. SEPs were elicited by electrical tail-base stimulation, and SI-SEPs and Vx-SEPs were recorded simultaneously. Responses to increasing stimulus intensity and stimulus frequency while awake and responses to bolus injection of fentanyl, thiopental, and ketamine were investigated. The SI-SEP positive component (P) occurring at 12 ms after stimulation (P12) showed a significantly lower intensity threshold and was significantly less affected by increasing stimulus frequency and by administration of the different drugs when compared to the Vx-SEP P15. The fact that a single stimulus modality results in different signal characteristics dependent on the recording site supports the view that different neural mechanisms involved in primary processing of somatosensory information are responsible for the generation of the SI-SEP P12 and Vx-SEP P15, respectively. This differentiation between SI-SEPs and Vx-SEPs potentially has distinct consequences using the SEP to evaluate nociception and analgesia in the rat model.