Objective: While psychiatric disorders have been recognized as a risk factor for COVID-19 outcomes, the impact of substance use disorders (SUD) on COVID-19 outcomes has not, to date, been examined in a systematic manner. We examined the association between SUD (cannabis, cocaine, alcohol, opioid, and benzodiazepine) as well as psychiatric diagnoses (schizophrenia, mood disorders, anxiety disorders) and COVID-19 outcomes in a large, retrospective cohort study. Methods: COVID-19–positive patients admitted to a large health care system in the US between January and December 2020 were included in this study. SUD and psychiatric diagnoses were identified from urine toxicology reports and ICD-10 diagnosis codes in the electronic medical record, respectively. Multivariable logistic regression was performed controlling for potential confounders such as age, race, sex, smoking status, and medical comorbidities. COVID-19–relevant outcomes included mortality, need for intensive care unit (ICU) admission, need for ventilatory support, length of hospitalization, and number of hospitalizations. Results: Among COVID-19 patients (N = 6,291), those with SUD were more likely to require ICU admission (adjusted odds ratio = 1.46, P = .003) and ventilatory support (AOR = 1.49, P = .01). The association between SUD and ICU admission was driven by alcohol use disorder (AUD), whereas that between SUD and ventilatory support was driven by both AUD and opioid use disorder (OUD). Patients with SUD were more likely to have a longer mean maximum length of hospitalization (11.32 vs 8.62 days, P < .0001) and a greater mean number of hospital admissions in 2020 (2.96 vs 2.33, P < .0001). These associations were significant for cannabis use disorder, AUD, OUD, and benzodiazepine use disorder. The association with greater number of admissions was also significant for cocaine use disorder. Patients with psychiatric diagnoses were also more likely to have a greater maximum length of hospitalization (11.93 vs 8.39 days, P < .0001) and hospital admissions (2.72 vs 2.31, P < .0001). These associations were significant for schizophrenia, mood disorders, and anxiety disorders. Conclusions: COVID-19 patients with SUD had greater likelihood of requiring critical interventions, such as ICU admission and ventilatory support. SUD and psychiatric diagnoses were also associated with a longer duration of hospitalization and greater number of hospital admissions. These findings identify COVID-19 patients with SUD and psychiatric comorbidities as a high-risk group.