A cluster randomized trial demonstrates that an electronic health record–integrated clinical decision support system increases naloxone prescriptions and opioid use disorder treatments in primary care but does not change diagnosis rates or MOUD adherence.