Eighth nerve action potential (AP) amplitudes and latencies and cochlear microphonic (CM) amplitudes were compared using tympanic and transtympanic electrocochleography (ECOG) in two patient groups. Tympanic ECOG was performed with a wick electrode placed on the tympanic membrane (TM). Transtympanic ECOG was performed with a needle electrode placed on the promontory of the anesthetized patient. Eighteen subjects were tested by tympanic ECOG as part of a preoperative assessment for either acoustic neuroma removal or transection of the vestibular portion of the eighth cranial nerve. Surgery occurred within 1 week of the preoperative evaluation. Intraoperative auditory monitoring was performed using transtympanic ECOG. Baseline recordings were compared to the preoperative tympanic ECOG data. Stimuli were condensation and rarefaction clicks and tone bursts, presented by an insert earphone. As expected, the two methods resulted in essentially identical response latencies and large amplitude differences, although the response amplitudes were extremely variable. The AP amplitude and the CM amplitude did not increase by the same factor with the transtympanic (TT) electrode compared to the tympanic electrode. On comparison of preoperative and intraoperative response amplitudes with regard to stimulus polarity, tympanic electrocochleography appears to be a useful method of gathering preliminary information on the status of the patient's auditory system. In this study, tympanic ECOG was found to have some predictive value when trying to ascertain the best intraoperative monitoring situation.