7 Sober Living Statistics: What the Research Shows

Here's something most articles won't tell you upfront: no one actually knows how many sober living homes exist in the United States. The Government Accountability Office looked into this and concluded that nationwide recovery-housing prevalence is unknown, because the organizations tracking residences only capture subsets of what's out there.¹ A follow-up GAO testimony reiterated the same point—available information covers only a portion of recovery homes, and states have built their own certification and monitoring systems at different speeds and with different definitions.²

Why does that matter for you? Because when you read a headline statistic claiming a specific national success rate or a precise count of residences, that number is almost certainly built on partial data. There's no federal registry. There's no mandatory licensing standard across all 50 states. Some homes are peer-run and never show up on a state list; others are certified through the National Alliance for Recovery Residences but operate alongside uncertified homes on the same block.

The most serious attempt to size the field came from a 2021 peer-reviewed study that combined data from multiple recovery-home networks and expert interviews. Researchers identified:

• 2,355 Oxford Houses
• 1,470 NARR-listed homes
• 1,125 homes in a SAMHSA database
• 12,993 listings from another database

—landing on a combined estimate of 17,943 recovery homes serving approximately 274,528 people in a given year.⁵

Treat that figure as a floor, not a ceiling. The authors themselves called it preliminary, warned that unlisted homes are missing from the count, and noted that listing quality varies widely. A home on a database isn't automatically a good home, and a home that isn't listed isn't automatically a bad one. The study's purpose was partly to make the case for a centralized, regularly updated national database—which still doesn't exist.

What you can take from this number: recovery housing is not a niche experiment. Hundreds of thousands of people in the United States spend part of their recovery year inside one. That's a meaningful amount of lived experience, even if the system around it hasn't been fully mapped yet. When you're choosing between homes in Portland, Seattle, or anywhere else, you're joining a large, uneven landscape—so your questions about staffing, structure, and peer culture carry real weight.

Here's a stat that reframes the conversation. In a national survey of people who had resolved an alcohol or drug problem, about 9% reported ever accessing recovery housing—while only 28% reported accessing formal treatment at all.⁸ Most people who get better don't pass through either system. The ones who do use recovery housing are a self-selected group: often leaving detox or treatment, often rebuilding housing stability, often trying something new after earlier attempts didn't hold.

You may be in that 9% because you've already done the hard work of detox and want structure before you go back to the house, apartment, or neighborhood where using felt automatic. Or you may be looking at it as a starting point because you don't have a safe place to land. SAMHSA frames recovery housing as substance-free housing centered on peer support and connections to services—a specific kind of environment, not a treatment program on its own.³

Choosing it isn't a sign you've failed at recovery. It's a sign you're taking the next 90 to 180 days seriously enough to change where you sleep.

Here's one of the clearest short-term pictures the research has produced. In a 2024 peer-reviewed study of people entering sober living houses, the share reporting any abstinence during the prior six months rose from 26.4% at baseline to 57.4% at the six-month follow-up.⁷ The percentage of days abstinent—a more granular measure than yes/no—climbed from 71.3% to 88.6% over that same window, and the average length of stay was 130.2 days.⁷

Read those numbers carefully before you let them do too much work. They describe a specific sample of people who chose sober living, measured over six months, with the researchers watching what changed. They don't tell you what would have happened without the residence, and they aren't a national success rate.

What they do suggest is a pattern you'll see repeated throughout the literature: people who stay in recovery housing for roughly four months tend to use less, and use less often, than they did when they walked in. If you're in the first weeks after detox or an inpatient stay and wondering whether another 90 to 180 days of structure is worth it, this is the kind of evidence that says it often is—for the people who stay.

The strongest comparative evidence in the field comes from a randomized study of Oxford House, a self-run, mutual-help recovery housing model. Researchers followed participants over 24 months after substance-use treatment and compared people assigned to Oxford House with people who received usual care. At the two-year mark, 31.3% of Oxford House participants reported substance use, versus 64.8% of the usual-care group.²² Employment ran 76.1% in the Oxford House group and 48.6% in the comparison group.²²

Those are big gaps, and they're among the most-cited numbers in recovery housing because they come from random assignment rather than self-selection. That matters: when people get to choose, you can't fully separate the effect of the house from the effect of being the kind of person who signs up.

One honest caveat. Oxford Houses are peer-run, democratically operated homes without on-site staff. They are not the same as a provider-operated sober living program or a clinically supervised residence in Portland or Seattle, and the results don't automatically transfer across models.²² Treat the 31.3% versus 64.8% substance-use gap as evidence that a specific, well-defined model of recovery housing works better than being discharged into whatever comes next—not as a universal claim about every residence with a sign out front.

In 2025, a systematic review pulled together the strongest comparative studies on recovery housing to answer a simple question: when researchers run controlled trials, does recovery housing beat usual care? The review identified five eligible studies—three randomized controlled trials and two quasi-experimental studies—with samples ranging from 150 to 470 people and follow-up periods of six to 24 months.²¹

The headline finding: recovery housing generally produced better outcomes than usual care or no intervention for abstinence, income, employment, criminal charges, and some incarceration measures. The reviewers characterized the overall evidence as moderate and called for more comparative trials.²¹

Two honest notes before you file that away as a win. First, five studies is a small base for a category that houses hundreds of thousands of people a year. Second, benefits weren't uniform across every population; findings were less consistent in samples with high proportions of formerly incarcerated women.²¹ The research supports recovery housing as a credible step after treatment for many people. It doesn't say every residence works equally well for everyone, which is why the questions you ask about a specific home—its structure, its peer culture, how long people actually stay—still matter.

Sober living isn't just about not using. For a lot of people, it's about having a key to a door that won't be taken away in two weeks. One of the clearest datasets on that question comes from a longitudinal study of 299 adults entering 20 sober living houses, which tracked housing status from the day they moved in through the 18-month follow-up.¹²

Here's what shifted over that year and a half:

• Homelessness among residents dropped from 16% at entry to 4%.
• Marginal housing—couches, motels, places that could end any day—fell from 66% to 46%.
• Stable housing climbed from 13% to 27%.¹²

Read those numbers for what they are. Most residents didn't walk straight from a sober living house into their own lease; the biggest shift was out of homelessness and into something less fragile. The share of people in genuinely stable housing roughly doubled, which is meaningful, but more than half were still in some form of transitional or marginal setup at 18 months.

If you're weighing whether a residence in Portland or elsewhere in the Pacific Northwest is worth the next few months, that's the honest expectation. You're trading an unstable floor for a steadier one, and giving yourself time to build toward the lease, the roommate situation, or the family reunification that comes after.

The same 300-person sober living study tracked more than housing. Over 18 months, researchers documented significant, sustained improvements in alcohol and drug use, employment, arrests, and psychiatric symptoms among residents entering two types of sober living houses.^{11, 13} A separate review of the social-model recovery literature summarized the same pattern across the field: residents showed gains in substance use, abstinence, psychiatric symptoms, employment, and arrests over that window.¹⁰

Those four outcomes tend to move together, and that's not a coincidence. When you're sleeping somewhere safe, you can hold a job. When you're holding a job, you're less likely to be arrested for the kinds of things people get arrested for when they're using and broke. When arrests drop and work picks up, the anxiety and depression that fed earlier use often ease too—not disappear, but ease.

None of this is automatic. The studies measured change among people who chose sober living and stayed, not a random cross-section of everyone who tried it. What the research does tell you: the direction of travel on employment, arrests, and mental health is real for residents who engage, and the effects were still visible 18 months after move-in.^{10, 11, 13} That's a long enough window to matter for a parole officer, a boss, or a judge watching your progress.

One number from the Oxford House randomized trial has stuck around because it answers a question families and funders ask directly: is this worth paying for? Over 24 months, researchers estimated a net benefit of approximately $29,000 per person for participants assigned to Oxford House, after accounting for differences in substance use, employment, illegal activity, and incarceration compared with usual care.²²

Keep the scope in mind. Oxford Houses are peer-run and democratically operated, so the economics look different from provider-staffed homes.²² SAMHSA has also flagged financing as a serious sustainability challenge for recovery residences generally, which is why rent, scholarships, and public funding vary so widely from one home to the next.⁴

Still, when a family asks whether a few months of rent in a sober living house pencils out against another relapse, the research gives you a credible answer: for one well-studied model, the math worked.