Medication-assisted treatment of opioid use disorder with physiological dependence at least doubles rates of opioid-abstinence outcomes in randomized, controlled trials comparing psychosocial treatment of opioid use disorder with medication versus with placebo or no medication. This article reviews the current evidence for medication-assisted treatment of opioid use disorder and also presents clinical practice imperatives for preventing opioid overdose and the transmission of infectious disease. The evidence strongly supports the use of agonist therapies to reduce opioid use and to retain patients in treatment, with methadone maintenance remaining the gold standard of care. Combined buprenorphine/naloxone, however, also demonstrates significant efficacy and favorable safety and tolerability in multiple populations, including youth and prescription opioid-dependent individuals, as does buprenorphine monotherapy in pregnant women. The evidence for antagonist therapies is weak. Oral naltrexone demonstrates poor adherence and increased mortality rates, although the early evidence looks more favorable for extended-release naltrexone, which has the advantages that it is not subject to misuse or diversion and that it does not present a risk of overdose on its own. Two perspectives-individualized treatment and population management-are presented for selecting among the three available Food and Drug Administration-approved maintenance therapies for opioid use disorder. The currently unmet challenges in treating opioid use disorder are discussed, as are the directions for future research.