Recovery housing, an abstinence-based living environment, is the most widely available form of substance use disorder (SUD) recovery support infrastructure. This systematic review characterized the randomized control trials (RCT) and quasi-experimental designs (QED) research on recovery housing. We conducted a search across PubMed, EMBASE, CINAHL, PsycINFO and CENTRAL published prior to February 2024. For inclusion, studies had to compare recovery housing alone to a non-recovery housing condition. Our search identified 5 eligible studies including 3 RCTs and 2 QEDs, across 11 reports. Participants Ns ranged from 150 to 470 and follow-up durations were 6-24 months. Recovery housing interventions performed better than continuing care as usual/no intervention on abstinence, income, employment, criminal charges and to a lesser extent incarceration. Recovery housing also performed better than comparative interventions delivered in other types of residential settings (e.g., therapeutic communities) on increasing alcohol abstinence and reducing days of substance use, while also increasing income and employment rates. An exception was in study samples that had high percentages of formerly incarcerated women (90% or more) where reduced substance use was the only benefit of recovery housing when compared to other types of residential interventions and was inconsistent when compared to continuing care as usual/no intervention. Moreover, recovery housing demonstrated higher cost effectiveness than continuing care as usual/no intervention and comparative interventions. Based on quantity, quality, and support for the service, the existing level of evidence for recovery housing is considered moderate. Expanding access to recovery housing may enhance outcomes for individuals with SUD, in general, while producing cost saving benefits, but given the small number of high quality studies additional comparative trials are needed. Also, future research should identify specific sub-groups who may or may not benefit from recovery housing interventions and why, so as to develop and test suitably augmented housing models or identify helpful alternatives.