ObjectiveTo test a method to measure the efficacy of active middle ear implants when coupled to the round window.MethodsData previously published in Koka et al. (Hear Res 2010;263:128–137) were used in this study. Simultaneous measurements of cochlear microphonics (CM) and stapes velocity in response to both acoustic stimulation (forward direction) and round window (RW) stimulation (reverse direction) with an active middle ear implant (AMEI) were made in seven ears in five chinchillas. For each stimulus frequency, the amplitude of the CM was measured separately as a function of intensity (dB SPL or dB mV). Equivalent vibrational input to the cochlea was determined by equating the acoustic and AMEI-generated CM amplitudes for a given intensity. In the condition of equivalent CM amplitude between acoustic and RW stimulation-generated output, we assume that the same vibrational input to the cochlea was present regardless of the route of stimulation.ResultsThe measured stapes velocities for equivalent CM output from the two types of input were not significantly different for low and medium frequencies (0.25–4 kHz); however, the velocities for AMEI-RW drive were significantly lower for higher frequencies (4–14 kHz). Thus, for RM stimulation with an AMEI, stapes velocities can underestimate the mechanical input to the cochlea by ~20 dB for frequencies greater than ~4 kHz.ConclusionsThis study confirms that stapes velocity (with the assumption of equivalent stapes velocity for forward and reverse stimulation) cannot be used as a proxy for effective input to the cochlea when it is stimulated in the reverse direction. Future research on application of intraoperative electrophysiological measurements during surgery (CM, compound action potential, or auditory brainstem response) for estimating efficacy and optimizing device coupling and performance is warranted.