1. We tested the coincidence, or cross-correlation, model of Jeffress, which proposes a neuronal mechanism for sensitivity to interaural time differences (ITDs) in low-frequency cells in the central nucleus of the inferior colliculus (ICC) of the cat. Different tokens of Gaussian noise stimuli were delivered to the two ears. We studied the neural responses to changes in ITDs of these stimuli and examined the manner in which the binaural cells responded to them. All of our results support the idea that the central binaural neurons perform an operation very similar to cross-correlation on the inputs arriving from each side. These inputs are transformed from the actual acoustic signal by the peripheral auditory system, and these transformations are reflected in the properties of the cross-correlations. 2. The responses to ITDs of identical broadband noise stimuli to the two ears varies cyclically as a function of ITD at a frequency close to the best frequency of the neuron. This cyclic response is a consequence of the narrowband filtering of the wideband acoustic signal by the auditory nerve fibers. To examine the effects of using stimuli to the two ears that were correlated to each other to different degrees, we generated pairs of noises. Each pair consisted of one standard noise, which was delivered to one ear, and a linear sum of two standard uncorrelated noises, which was delivered to the other ear. The responses of 34 neurons in the ICC to ITDs of noises with variable interaural coherence were examined. When partially correlated noises were delivered, there was a positive and approximately linear relationship between the degree of modulation of the response as a function of ITD and interaural coherence. The degree of modulation was measured by the synchronization coefficient, or vector strength, over one period of the ITD curve. 3. We examined the effects of altering the interaural phase relationships of the input noise stimuli. The phase of the noise stimuli was changed by digitally filtering the standard noise so that only a phase delay was imposed. The responses to ITDs with differing interaural phase relationships were then studied by delivering a phase-shifted noise to one ear and the standard noise to the other. The ITD curves in response to phase-shifted noise were shifted by about the same amount as the shift of the stimulus; the shift of the response was measured with respect to the case with identical noises to the two ears.(ABSTRACT TRUNCATED AT 400 WORDS)