Cognitive deficits are a core feature of schizophrenia. There is ongoing debate on whether cognition is affected by antipsychotic drugs (APDs). This study examined the effect of long-term treatment with APDs on cognition in schizophrenia. Cognitive function was assessed in 418 patients with schizophrenia on long-term treatment with APDs (215 on clozapine, 91 on risperidone and 112 on typical APDs) and 159 healthy controls using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Schizophrenia symptomatology was assessed using the Positive and Negative Syndrome Scale (PANSS). We found that cognitive test scores were significantly lower in all patients compared with the healthy controls on almost all of the total and subscores of RBANS (all P<0.001), except for the visuospatial/constructional index. Individuals taking clozapine showed worse immediate and delayed memory performance than those taking typical APDs (all P<0.01). Moreover, individuals taking clozapine showed better language performance than those taking risperidone (P<0.01). Immediate memory and delayed memory were modestly correlated with the types of APDs and the PANSS negative scores. Our results show that individuals taking clozapine performed worse in immediate and delayed memory than those taking typical APDs, but exemplified better language performance than those taking risperidone.