OBJECTIVE The aim of this study was to examine selective attention in patients with chronic and refractory schizophrenia who had been exposed for six months to atypical neuroleptic medications: risperidone or clozapine. METHODS 17 patients satisfying DSM III-R criteria for schizophrenia were assessed according to BPRS and PANSS and abnormal involuntary movements to ESRS. Selective attention tasks were performed before treatment with risperidone or clozapine and at two times during the treatment (6 weeks, T1, and 24 weeks, T2). Patients' performance data were compared to data from a group of general population at T1. Selective attention refers to the ability to discriminate relevant information from irrelevant one. This was measured by a visual search task. Subjects had to search for a target specified by a conjunction of features (color and shape). The target was a black X, while the distracters were white X's, black O's and white O's. The stimuli were displayed on a Macintosh SE computer. A two-button response box was used for response production and the experiment was run in a dimly lit room. A white-fixation stimulus was shown at the center of the display screen between trials. The number of stimuli displayed on a single trial was 1, 4, 7 or 10. The median RTs and error rates of subjects were computed for each factor (target presence and number of stimuli). RESULTS A Group X Number of items X Presence of target ANOVA applied on median correct RTs revealed a significant Group X Presence interaction [F(1,176) = 60.433, p < .0001]. Performances improved with the time (F2, p < .01). Correlations were found between positive score on PANSS and performance on selective attention (r39 = -.391). CONCLUSIONS Atypical neuroleptic do not have a deleterious effect on selective attention but a favorable effect on the schizophrenic patients' performance.