Sexual minority populations experience higher rates of substance use and related problems, but little is known about their specific involvement in driving under the influence (DUI) of alcohol (DUIA) and cannabis (DUIC) incidents. Using data from the 2016 to 2019 National Survey on Drug Use and Health, we used logistic regression models to estimate the interactive effects of sexual identity, race/ethnicity, and gender on past-year DUIA among adults who used alcohol and DUIC among adults who used cannabis, accounting for covariates. Using model estimates and linear combinations, we calculated the predicted probabilities of each outcome and compared sexual identity differences within and across race/ethnicity and gender. With few exceptions, the predicted probabilities of DUIA and DUIC were significantly higher among sexual minority women than heterosexual women of similar race/ethnicity. The results were more variable among men with the probabilities of DUIA and DUIC being significantly higher for some groups of sexual minority men and some groups having probabilities equal to or lower than similar heterosexual men. Some of the largest sexual minority gaps in DUIA and DUIC were observed among Hispanic and Other lesbian women and Black gay men. Sexual minority individuals are more likely to report DUI than their heterosexual counterparts; however, the risk of DUI among sexual minority populations varies by racial/ethnic and gender subgroup. Our findings indicate the importance of applying an intersectional framework when addressing substance-use-related disparities and when designing effective DUI prevention interventions for sexual minority populations.