Panic attacks are subjectively reported by patients as a cluster of autonomic nervous symptoms. Taken together with the evidence from pharmacological studies focusing on the locus ceruleus and the dorsal raphe nuclei, the hypothesis of brainstem involvement in panic attacks is of major interest. Functional evaluation of the brainstem is carried out using electrophysiological recordings, such as the brainstem evoked potentials. We have investigated the pathophysiology of these parameters in 16 patients fulfilling the DSM III-R criteria for panic disorder (8 females and 8 males) in comparison to a group of 10 normal controls. The patients were found to have two electrophysiological variables significantly different from controls: (a) N3 latency time was reduced (p < 0.05) and (b) the N3–5 interval was prolonged (p < 0.05). The N3 latency period significantly correlated with the Hamilton Anxiety Scale scores (p < 0.005). N3 is a peak reflecting pontine activation, where the locus ceruleus is located. The N3–5 interval reflects evoked potential passage from the pons to the midbrain. In conclusion, our results support previous studies indicating a dysfunction of the locus ceruleus possibly causing disruption of pons-mid-brain transmission in patients suffering from panic disor