OBJECTIVE:This study used statewide administrative data sets to estimate the prevalence of tobacco use among persons with mental illnesses who were accessing public-sector mental health care in Colorado and to determine the relationships between tobacco use and primary diagnosis and alcohol and drug use. METHODS:This study utilized the Colorado Client Assessment Record to examine predictors of tobacco use among 111,984 persons with mental illnesses who were receiving services in the public mental health system. RESULTS:Thirty-nine percent of the sample (N=43,508) used tobacco. Multiple logistic regression analysis found that schizophrenia, schizoaffective disorder, and bipolar disorder (p<.001 for all), and depression or dysthymia (p<.01) were associated with greater tobacco use than other diagnoses. Significant differences in tobacco use existed across gender, age group, race or ethnicity, and substance use categories. CONCLUSIONS:Findings suggest that an administrative database is a low-burden means of identifying persons at high risk of tobacco use to inform resource allocation.