Male schizophrenia patients are known to have a heavier smoking pattern compared with the general population. However, the mechanism for this association is not known, though hypothesis that smoking could alleviate symptomatology of schizophrenia and reduce side effects of antipsychotics has been suggested. The aims of this study were to validate the heavier smoking pattern among male schizophrenia patients and to investigate the possible mechanisms for the association. To enhance the reliability of the study, we recruited two large independent samples with 604 and 535 male Chinese schizophrenia patients, and compared their smoking pattern with that of 535 healthy male controls recruited from general population. Validated multiple indicators and multiple causes structure equation model and regression models were used to investigate the association of smoking with factors of schizophrenia symptomatology and with the usage of antipsychotics and their extra-pyramidal side effects (EPS). Schizophrenia patients had significantly heavier smoking pattern compared with healthy controls in our sample (42.4% vs. 16.8%, p<0.001 for current smoking prevalence; 23.5% vs. 43.3%, p<0.001 for smoking cessation rate; 24.5% vs. 3.0%, p<0.001 for heavy smoker proportion). Their smoking status was also found to be consistently and significantly associated with reduced negative factor scores for schizophrenia symptomatology (β = -0.123, p = 0.051 for sample-A; β = -0.103, p = 0.035 for sample-B; β = -0.082, p = 0.017 for the combined sample). However, no significant association was found between smoking and antipsychotics usage or risk of EPS. These results support that smoking is associated with improved negative symptoms, which could account for the heavier smoking pattern among schizophrenia patients.