Registration of cavernous electrical activity (single potential analysis of cavernous electrical activity) was recently suggested for the diagnosis of autonomic cavernous dysfunction. For validation of this method the effect of sacral neurostimulation on cavernous electrical activity was examined. In 3 male patients with a complete spinal cord lesion (T3, T4 and T12, respectively), deafferentation was done at the S2 to S5 levels. Around the anterior roots of S2 to S5 electrodes for neurostimulation were placed. Cavernous electrical activity was recorded with an intracavernous needle electrode and with surface electrodes in the flaccid as well as in the erect states induced by neurostimulation (at 7, 8, 12, 18, 20, 30 and 45 Hz., 30 v. and 0.4 msec). In all patients similar potentials compared to the normal values, as well as additional pathological potentials were recorded during flaccidity. During neurostimulation all patients achieved full erection with no or markedly decreased cavernous electrical activity to frequencies of 12, 18, 20 and 30 Hz., while to 7, 8 and, more pronounced, 45 Hz. only partial erection with ongoing cavernous electrical activity was found. Our study strongly suggests that cavernous electrical activity and, subsequently, the cavernous smooth muscle tone are dependent on autonomic input. This finding supports the hypothesis that single potential analysis of cavernous electrical activity may be valid in the diagnosis of cavernous autonomic dysfunction. Furthermore, our results suggest a possible role for single potential analysis of cavernous electrical activity in the fine tuning of erection inducing neurostimulators.