Transcranial magnetic motor evoked potentials (TMMEP) were recorded from the extensor carpi radialis muscle of the thoracic limb and from the cranial tibial muscle of the pelvic limb in anesthetized dogs. All dogs were premedicated with droperidol and fentanyl. Anesthesia was induced and maintained with either propofol, sufentanil and midazolam, thiopental, diazepam and ketamine, halothane, or nitrous oxide, the latter both with and without fentanyl. The effect of different doses of these anesthetics upon onset latencies and peak-to-peak amplitudes of the TMMEP was evaluated. Anesthetics produced a dose-dependent suppression of the TMMEP: latencies increased, amplitudes decreased, reproducibility became poorer and the number of unsuccessful recordings increased with increasing depth of anesthesia. Anesthesia using narcotics (sufentanil and midazolam, fentanyl and nitrous oxide) was considered superior to the other types of anesthesia because they produced the least suppression of the TMMEP and allowed TMMEP recording in all dogs, even at surgical planes of anesthesia.