Sexual-identity disparities in substance use among U.S. veterans, and whether mental-health treatment mitigates risk for those with depression, remain under-examined. Using data on veterans from the 2021–2023 National Survey on Drug Use and Health (NSDUH; N = 7,212), disparities were estimated in past-30-day nicotine, marijuana, binge drinking, and polysubstance use, as well as severe psychological distress (K6≥13) and past-year suicidal ideation. Guided by a biosocial/minority-stress framework, multiple imputation was applied (m = 20) and survey-weighted logistic regression adjusting for age, year, race/ethnicity, sex, education, metro status, insurance, marital status, employment, and income; among veterans with a past-year major depressive episode (MDE), interactions were tested between sexual identity and (a) depression-related clinical contact (DRC) and (b) prescription medication for depressive feelings. Bisexual veterans showed the highest prevalence of marijuana (33.5%) and polysubstance use (30.6%), exceeding that of heterosexual (11.8%, 14.9%) and gay/lesbian veterans (24.0%, 18.8%). Models restricted to veterans with MDE, past-year DRC (DRC defined as any visit or conversation with a health professional about depressive feelings) moderated risk for gay/lesbian veterans, with DRC associated with lower odds of binge drinking and polysubstance use; prescription medication showed a similar moderating pattern for nicotine and polysubstance outcomes. Findings for severe psychological distress and suicidal ideation were mixed and consistent with confounding by indication. Results should be interpreted cautiously given the cross-sectional data, self-report, small sexual-minority subgroups, and non-aligned recall windows (past-year mental health/treatment vs past-30-day substance use). Overall, sexual-identity disparities in substance use are evident, with bisexual veterans bearing the greatest burden, and engagement in DRC and medication among veterans with MDE, particularly gay/lesbian veterans, showing associations consistent with a buffering effect of affirming care. Longitudinal and qualitative studies are needed to test causal pathways and to illuminate lived experiences, and policy/clinical efforts should expand culturally competent, integrated services and routine SOGI data collection to monitor and reduce inequities.