The responses of neurons in the posteroventral (PVCN) and dorsal (DCN) cochlear nucleus of the unanesthetized cat were determined for both long and short tones. These results were compared with recent studies in the barbiturate-anesthetized cat conducted in the same laboratory using similar stimuli and analysis programs. Every response pattern (poststimulus time histogram to short tones), which has been observed in previous studies using anesthetized animals, was also observed without anesthetic. The converse was also true: no novel response patterns were observed in the unanesthetized cat. This was also true for interval histogram, response area, isorate curve, and frequency sweep data. Some neurons were difficult to classify into existing descriptions of cochlear nucleus response patterns. For example: primary-like, onset, pauser, and buildup response patterns could also show chopper-like properties; onset-inhibitory, pauser, and buildup neurons appeared to form a response continuum rather than exist as separate response categories; and onset neurons with low characteristic frequencies (CFs) often showed sustained and strongly phase-locked responses below approximately 1,000 Hz. In addition, single neurons often showed more than one response pattern depending on the intensity and frequency of the acoustic stimulus. These ambiguities were also observed under anesthetic. Onset neurons within the PVCN appear to be well suited for the encoding of temporal and intensity information. At low stimulus frequencies they often respond to every cycle of a pure tone stimulus and exhibit the highest degree of phase-locking in the cochlear nucleus. The dynamic ranges associated with many onset neurons can exceed 80 dB compared with the 30- to 40-dB dynamic ranges associated with most other cochlear nucleus neurons. Onset neurons show a similar range of activities in the anesthetized cat. Neurons in the DCN have response properties that are more complex than those seen in the PVCN. Response patterns can change from sustained excitation to complete inhibition and are more often nonmonotonic near CF. DCN neurons can show well-defined tuning in the frequency domain and may be used to encode spectral information, but appear to be poorly suited for encoding temporal or intensity information as they are weakly phase-locked and have relatively small dynamic ranges. When DCN neurons "chop" they usually do so more slowly than do PVCN neurons. DCN neurons recorded in the anesthetized cat behave similarly. The relative frequency of a particular response pattern did vary with anesthetic state.(ABSTRACT TRUNCATED AT 400 WORDS)