Dichotic stop-consonant-vowel identification was investigated in two experiments using two groups of learning-disabled children demonstrating clinical manifestations of auditory-processing disorders, and two groups of matched, control subjects. Two-item, forced-choice paradigms were used in both experiments. Overall (total) dichotic performance for the two learning-disabled groups was significantly lower than that of the control subjects in all dichotic conditions. This lower performance was attributable to the number of trials in which both stimulus items were correctly identified. Analysis of trials in which only one response was correct showed no differences between the groups in terms of magnitude or direction of ear-advantage (right). In conditions where stimulus onsets were separated by 30, 90, and 150 msec, analysis of one-correct trials demonstrated more accurate identification of the temporally lagging stimulus for all subjects. However, as the onset-time separation increased, the control group’s identification of leading and lagging items approached equality. The learning-disabled group, on the other hand, showed little increase in identification of temporally leading items even when stimuli were separated by 150 msec. These data suggest learning-disabled children with auditory-processing disorders may have a reduced temporal efficiency in processing rapidly varying acoustic patterns associated with stop-consonants that is observable when speech perceptual mechanisms are stressed through dichotic competition.