Schizophrenic patients have higher smoking rates than the general population. Studies show that smoking may be a form of self-medication in an attempt to alleviate cognitive deficits in schizophrenic patients of European background. This study examined the relationships between smoking and cognitive deficits in Chinese schizophrenic patients, which have previously received little systemic study. We recruited 580 male chronic patients meeting DSM-IV criteria for schizophrenia and 175 male control subjects who were matched on age and education. The subjects completed a detailed cigarette smoking questionnaire, the Fagerstrom Test for Nicotine Dependence (FTND), and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Patients also were rated on the Positive and Negative Symptom Scale (PANSS), the Simpson and Angus Extrapyramidal Symptom Rating Scale (SAES), and the Abnormal Involuntary Movement Scale (AIMS). All five RBANS subscales except for the Visuospatial/Constructional index showed significantly lower cognitive performance for schizophrenics than normal controls. The schizophrenic smokers scored lower than the schizophrenic non-smokers on the RBANS total score and the Visuospatial/Constructional and Immediate Memory indices. Similarly, the control smokers scored lower than the control non-smokers on the RBANS total score and the Immediate Memory index . Also, the schizophrenic smokers consistently performed the poorest on the cognitive domains of the RBANS. Among the schizophrenic patients, smokers displayed significantly fewer negative symptoms than non-smokers. Using multivariate regression analysis the following variables were independently associated with the RBANS total score: years of education, PANSS negative symptom score, age at schizophrenia onset, and number of hospitalizations. Our results show that smoking is associated with significant cognitive impairment in both schizophrenic patients and normal controls, but the smokers with schizophrenia had a reduced level of negative symptoms, suggesting that the benefits of smoking for those with schizophrenia may be limited to certain aspects of a given clinical phenotype.