Previous studies of the auditory P300 event-related potential (ERP) have reported smaller amplitudes in chronic schizophrenics but similar consistencies have not been observed with visual P300s. This study examined P300s in symptomatically stable, medicated, chronic schizophrenics (n = 14) and normal controls (n = 14) performing a visual continuous performance task utilizing degraded stimuli to burden encoding processes. Performance analysis found slower response times, fewer target detections and more false alarms in patients than in controls. Analysis of ERPs showed P300 amplitudes of schizophrenics to be significantly smaller than those of controls and, unlike controls, schizophrenics failed to exhibit significant target vs. non-target P300 amplitude differences. Discriminant analysis indicated target and non-target midline (F-Z, C-Z, P-Z) P300 amplitudes together correctly classified all patients and controls. Exploratory topographic analysis indicated that P300 amplitudes were not asymmetrical in patients, as has been observed with auditory P300s, and, unlike the performance measures, the P300s did not correlate with the patient's positive or negative symptom ratings. The implications of these findings are described in relation to attentional disturbances and trait versus state issues in schizophrenics. (C) 1999 Elsevier Science B.V. All rights reserved.