Aims To examine the associations between youth poly‐tobacco use and substance use disorders. Design Analysis of data from the 2007–11 US National Survey on Drug Use and Health. Setting Randomly selected, household‐dwelling adolescents from the non‐institutionalized, civilian population of the United States. Participants A total of 91 152 adolescents (aged 12–17 years). Methods Logistic regression models were used to examine the associations between type of tobacco user (non‐user, users of alternative tobacco products only, users of cigarettes only and users of cigarettes plus alternative tobacco products) with past year alcohol, marijuana or other illicit drug use disorders, adjusting for demographic and social variables. Findings Compared with non‐users of tobacco, the greatest risk for substance use disorders was among users of cigarettes plus alternative tobacco products [alcohol disorder adjusted odds ratio (aOR) = 18.3, 95% confidence interval (CI) = 16.2–20.6; marijuana disorder aOR = 37.2, 95% CI = 32.5–42.7; other drug disorder aOR = 18.4, 95% CI = 15.4–21.8], followed by users of cigarettes only (alcohol disorder aOR = 9.6, 95% CI = 8.8–10.6; marijuana disorder aOR = 20.4, 95% CI = 18.1–23.0; other drug disorder aOR = 9.4, 95% CI = 7.8–11.4), then users of alternative tobacco products only (alcohol disorder aOR = 8.1, 95% CI = 6.7–9.6; marijuana disorder aOR = 9.2, 95% CI = 7.5–11.4; other drug disorder aOR = 3.2, 95% CI = 2.4–4.3). Conclusions Tobacco use in adolescence is associated with higher rates of substance use disorders across all tobacco users, especially among those who use cigarettes plus other tobacco products.