Objectives: To determine if Electrode Voltage (EV) measurements are potentially suitable as a test for detecting extra-cochlear electrodes in cochlear implants (CIs) and to analyse voltages developed in different stimulation modes. Methods: EV measurements were made using surface electrodes in live mode without averaging in 17 adult cochlear implant (CI) users (18 ears). The effect of stimulation level, the position of the active recording electrode, repeatability and the position of the implant's active electrode were investigated. For Nucleus recipients, measurements were made in different stimulation modes. Results/Discussion: Recordings made in monopolar mode showed good repeatability when the active recording electrode was placed on the ipsilateral earlobe; voltages increased linearly with stimulation level as expected. Basal electrode measurements differed greatly between fully inserted devices with all electrodes activated, partially inserted/migrated devices and fully inserted devices with deactivated basal electrodes [χ 2 (2)=10.2, p<0.05 for the most basal electrode]. Analysis of voltages measured in different stimulation modes found that EVs could be separated into two components, associated with the CI's active and return electrodes respectively. EVs for electrodes on the array were much smaller than those for monopolar return electrodes for fully inserted devices, but were abnormally large for one participant with extra-cochlear electrodes. We argue that fibrosis around the electrode array facilitated current flow across the round window in this case. Conclusion: The test appears to be a viable approach to detect electrode migration and extracochlear electrodes in adult CI users and may also be sensitive to discomfort caused by current leakage from the basal end of the cochlea.