To predict the possible benefit of a cochlear implant to a patient, the clinician requires a reliable objective method to monitor the presence of surviving fibres in the 8th nerve. The subjective promontory test and the evaluation of electrically evoked brainstem potentials are not adequate. An electrical stimulus artefact interferes badly with the potential recorded, but with electrical sinusoidal stimulation this stimulus artefact can be suppressed. The highly synchronous excitation of the auditory nerve fibres results in compound action potentials that can be recorded from the round window. The present animal experiments show that the method proposed is effective. Clinical tests are suggested.