AbstractBackground and aimsTobacco use may be a risk factor for suicide mortality; however, prior research has produced equivocal findings and has been limited by relatively small sample sizes to study the rare event of suicide, as well as a lack of adjustment for other important factors, including psychiatric illness. We estimate the predictive association between tobacco use disorder and the risk of suicide mortality, adjusting for other important variables.DesignA prospective cohort study.SettingTheUnitedStatesVeteransHealthAdministration (VHA).ParticipantsAll individuals who receivedVHAservices in fiscal year (FY) 2005 and were alive at the start of FY 2006 (n = 4 863 086).MeasurementsTobacco use disorder was assessed via FYs 2004–05VHANationalPatientCareDatabase records. The outcome of suicide mortality was assessed during the follow‐up interval from the beginning of FY 2006 to the end of FY 2008 usingNationalDeathIndex records.FindingsOf the 4 863 086 individuals in the study, 4823 died by suicide during the follow‐up interval. In the unadjusted model, tobacco use disorder was associated with an increased risk of suicide [hazard ratio (HR) = 1.88; 95% confidence interval (CI) = 1.76, 2.02]. After adjustment for model covariates, the association remained statistically significant, although attenuated (HR = 1.36, 95% CI = 1.27, 1.46).ConclusionsTobacco use disorder may confer a modest excess risk of death by suicide. Psychiatric disorders may partially explain the relationship between tobacco use disorder and suicide.