Auditory brainstem responses (ABRs) were recorded using pseudo-random pulse trains called maximum-length sequences (MLSs), which allow high stimulus rates and simultaneous recording from both ears, and were compared to ABRs obtained by conventional averaging. In normally hearing subjects, ABRs by MLSs produced the same waves as conventional ABRs, although wave latencies increased and amplitudes decreased. In normal as in sensorineural ears, MLS-ABR thresholds were similar to conventional ABR thresholds (except when binaural recordings were made) and were correlated with high-frequency audiometric thresholds. Binaural MLS-ABR recordings represent the individual response of each ear, as shown by recordings of monaurally deaf subjects.