This study examined the utility of the clusters found in Poland et al. (2000) using linear and logistic regression. As well, this study aimed to replicate the findings of Poland et al. (2000) which identified heterogeneity among Ontarians with respect to their attitudes and behaviours concerning smoking and smoking restrictions. They identified three groups of smokers (adamant, easy going, reluctant) and non-smokers (adamant, unempowered, laissez-faire).Analyses revealed that cluster membership significantly predicted support for tobacco control policies and smoking restrictions among both smokers and non-smokers. Using cluster analysis on a short-form of the variables used in the previous study, it was found that smokers and non-smokers do cluster into subgroups based on attitudes regarding smoking and smoking restrictions. However, this study found the same clusters in different proportions than Poland et al. (2000). It is now evident that the clusters found in Poland et al. (2000) can be used to predict support for tobacco control policies and smoking restrictions.