A prior study in a US state hospital suggested that schizophrenia is more closely associated with tobacco smoking when compared with other severe mental illnesses. This second study, in a neighborhood hospital, tries to (1) replicate that schizophrenia is associated with smoking and heavy smoking, and (2) rule out that this relationship is explained by substance use. The methodology was very similar to the first study. The sample included 588 inpatients. Logistic regression was used to develop models of variables associated with smoking or heavy smoking. The frequency of current smoking for the total, schizophrenic and non-schizophrenic samples were respectively 71, 75, and 55%. The sequence of frequencies from the highest to lowest was the same as in the first study: male schizophrenic patients, male non-schizophrenic patients, female schizophrenic patients and female non-schizophrenic patients. This second study consistently replicated the relationship between schizophrenia and smoking (after correcting for other variables) including history of alcohol and drug abuse or dependence. Only one of two definitions of heavy smoking showed a significant association with schizophrenia. In summary, these two studies suggest that schizophrenia is strongly associated with smoking. Neither substance use, antipsychotics, nor institutionalism can explain this relationship.