The high reliability with which the waves I through V can be identified in auditory brain stem potential (BAEP) recordings suggests the possibility of using the computer to identify the peaks. This paper examines the application of a smoothing algorithm in which detection of zero crossings in the derivative array of the averaged potentials is used to identify waves I--V in clinical BAEP recordings. A comparison of computer wave identification with human evaluation is presented. The results suggest that computer evaluation cannot entirely replace 'manual' evaluation of the waves, but can expedite analysis and alert the clinician to potentially abnormal responses.