This study aimed to classify the responses of single units in the auditory midbrain to acoustic stimuli presented in the free field in order to characterize those units likely to have a role in sound localization in the horizontal plane. The responses of 131 single units in the inferior colliculus of the cat and the brush-tailed possum were studied using tone and noise-burst stimuli presented from a speaker capable of movement at any point along a plane 10 degrees above the horizontal plane. Speaker positions along this plane are referred to as speaker azimuths; those on the same side as the recorded inferior colliculus as ipsilateral, and on the opposite side as contralateral, azimuths. For each unit, spike counts were measured as a function of azimuth either at the best frequency (BF) or using noise bursts. These functions are referred to as azimuth functions and were usually measured for at least two intensities, between 10 and 70 dB above threshold. The recording sites of most units were identified histologically with the aid of microlesions and were related to the major subdivisions of the inferior colliculus: the central nucleus (ICC), the lateral part of the external nucleus (ICX), and the rostroventral process (R-ICX). Two units were located in the pericentral nucleus and two in the dorsal nucleus of the lateral lemniscus. Two major classes of neuron were identified: omnidirectional and directionally sensitive. Omnidirectional units exhibited azimuth functions that were either flat or that declined gradually at progressively ipsilateral azimuths. For the latter units, discharge rates at all points monotonically increased with stimulus intensity. There was no indication, for either type of omnidirectional unit, of significant binaural interaction. A good correlation was found between the summed proportions of excitatory-excitatory (EE) and monaural (EO) units observed in dichotic studies (46-55%) and the proportion of omnidirectional units in the present study (47%). A subgroup of directionally sensitive units (36% of the total) displayed azimuth functions for which the azimuthal position of the discharge border or peak firing azimuth remained essentially unaltered over a range of stimulus intensities. These azimuth-selective units are likely to have a role in the detection of the location of stimuli in the horizontal plane and appear to include units that would be considered excitatory-inhibitory (EI) or delay sensitive in dichotic studies. The azimuths over which directionally sensitive units showed their marked directional effects were influenced by the position of the contralateral pinna.(ABSTRACT TRUNCATED AT 400 WORDS)