The organization and connections of auditory cortex in owl monkeys, Aotus trivirgatus, were investigated by combining microelectrode mapping methods with studies of architecture and connections in the same animals. In most experiments, portions of auditory cortex were first explored with microelectrodes, neurons were characterized as responsive or not to auditory stimuli, and best frequencies were determined whenever possible. Most recordings were in cortex previously designated as primary (A-I) and rostral (R) auditory fields (Imig et al. J Comp Neurol 171:111, '77) and in a newly defined rostrotemporal field (RT) located rostral to R. Injections of wheat germ agglutinin-horseradish peroxidase (WGA-HRP) and fluorescent tracers were placed in electrophysiologically identified locations of A-I, R, and RT; the posterolateral (PL) and anterolateral (AL) divisions of a narrow belt of auditory cortex lateral and adjacent to A-I and R; cortex of the superior temporal gyrus lateral and rostrolateral to PL and AL; and regions of prefrontal cortex that receive inputs from auditory cortex. There were several major findings: 1. Best frequencies were most clearly determined for neurons within a densely myelinated strip of cortex on the lower bank and lip of the lateral sulcus. We divided this strip into three fields, A-I, R, and RT, although an alternative interpretation that A-I and R are parts of a single field remains tenable. In some cases, isofrequency contours appeared to continue uninterrupted across fields A-I and R, with lower frequencies represented laterally and higher frequencies represented deeper in the sulcus. In other cases, there was a tendency for high frequencies to be represented caudally and medially, and low frequencies laterally in A-I and rostrally in R, with partial discontinuity in the isofrequency contours. A reversal of the tonotopic gradient appeared in RT with a common low-frequency representation at the caudal border with R, and progressively higher frequencies encountered rostrally. Of the three fields, A-I appears slightly more myelinated than R, and RT slightly less than R. The distinctiveness of the three fields is further demonstrated by the patterns of connections. In particular, A-I and RT are both interconnected with R, but not with each other. Connections between A-I and R are between tonotopically matched locations. 2. A narrow 2-3 mm wide band of cortex lateral to A-I, R, and RT was also responsive to auditory stimuli, but typically neurons were more difficult to activate, and best frequencies were more difficult to determine. No distinctions in myeloarchitecture or CO activity were obvious.(ABSTRACT TRUNCATED AT 400 WORDS)