Acoustic evoked potentials to meaningful words were recorded in healthy, aphasic, and right hemisphere-damaged subjects under four conditions: monaural left and right, binaural, and dichotic stimulation. Four major findings emerged. First, healthy and brain-damaged subjects differed in amplitude and latency values of the N1 and P2 components. In healthy subjects N1 was greater and P2 smaller than in aphasics. Both components peaked earlier inpatients than in normals. Second, evoked potentials of healthy subjects showed a late sustained component which was decreased in aphasics. Third, the latencies of P1 and N1 as well as the amplitude of N1 showed a "pathway effect", i.e. shorter latency and greater amplitude to contralateral stimulation. Fourth, under the dichotic condition, P1 and N1 peaked earlier over the left hemisphere. The N1 amplitude behaved differently in the three groups depending upon stimulating conditions. It is suggested that these differences reflect linguistic coding and related attentional processes in patients and normals.