San Francisco – A new study from the UCSF Benioff Homelessness and Housing Initiative (BHHI) is reshaping the public narrative around homelessness and drug use in California. Contrary to common perceptions that homelessness is primarily driven by substance abuse, the comprehensive research found that just 37% of unhoused Californians reported regular use of illicit drugs.
Published in the medical journal JAMA, the report—Illicit Substance Use and Treatment Access Among Adults Experiencing Homelessness—offers one of the most in-depth looks at substance use and treatment barriers among California’s homeless population in decades.
The findings expose widespread misconceptions, while also highlighting the challenges facing those seeking help.
The study, based on 3,200 surveys and 365 in-depth interviews conducted in 2023, offers the most representative sample of California’s homeless population since the 1990s. Lead author Dr. Margot Kushel, director of BHHI, said the findings underscore the need for solutions grounded in housing and treatment—not stigma.
Among the most striking results:
- Regular drug use was reported by 37% of respondents, defined as using illicit drugs at least three times a week in the past six months.
- Lifetime drug use was much higher, with 65% reporting regular use at some point in their lives.
- New drug use began for 27% after becoming homeless, suggesting homelessness itself is a driving factor for some.
- A quarter of respondents said they had never used drugs at all.
- Treatment access remains abysmally low. Only 7% of respondents with any lifetime use were receiving treatment. Meanwhile, 20% sought help but couldn’t get it.
“One of the most poignant findings was that one in five told us they are actively seeking treatment and couldn’t get it,” Kushel said. “It shouldn’t be this hard. You shouldn’t have to scrounge for money and be put on a waiting list.”
The study challenges simplistic narratives that treat homelessness and drug use as interchangeable. While drug use is higher among homeless people than the general population, it is neither universal nor always causal.
Importantly, 35% of those surveyed said their drug use had decreased after becoming homeless. Some cited fear of losing their children, others simply said, “I just reached that point.” This finding debunks the myth that homelessness inevitably worsens substance use.
Where drug use does persist, it is often a survival mechanism. Methamphetamine use was most common, with respondents explaining that it helps them stay awake and alert—protection against assault, theft, or trauma in encampments.
“People are using it as a coping tool—to stay safe, to stay awake, to make it all go away,” Kushel explained.
Only 10% of respondents said they were regularly using opioids, often mixed with meth. Yet, the danger of overdose remains high. Nearly 20% of respondents reported experiencing an overdose in their lifetime, and 10% had overdosed during their current episode of homelessness.
Alarmingly, just a quarter of respondents had access to naloxone, the life-saving opioid overdose reversal drug. Kushel stressed that it should be far more widely available. “You can’t wait for a first responder,” she said. “People are seeing overdoses every day.”
Despite public attention on drug use and homelessness, the study reveals how inaccessible treatment remains. Of those reporting substance use, the vast majority were not receiving care—and many who tried were turned away.
“I’m showing up where they tell me, calling the numbers, and I’m still being told no,” one respondent shared.
Kushel called the gap “unacceptable” and blamed chronic underinvestment in treatment infrastructure. “We cannot talk about drug use without talking about the fact that people want help and can’t get it,” she said.
Ultimately, the researchers argue that housing is the most effective solution to both homelessness and substance use. “Housing is treatment,” Kushel said. “We know that without the stability of a home, it is incredibly difficult to address mental health or substance use challenges.”
Still, until enough housing is available, the report urges expanded access to street-based services like methadone or other opioid treatment medications, and more residential programs that lead directly to permanent housing.
Kushel also stressed that relapse must be understood as part of the recovery process. “It’s very common and part of the natural history of substance use disorders,” she said. Programs must support people through relapse—not push them back into homelessness.
The BHHI study arrives at a pivotal time as California grapples with how to address its homelessness crisis. Too often, drug use is cited as justification for punitive policies—forced removals of encampments, criminalization, or forced treatment models.
This research makes clear that the problem is not as simple as removing “drug users” from public spaces. With 63% of unhoused Californians not regularly using drugs, such policies risk punishing the majority who are sober or struggling with barriers far beyond substance use.
Moreover, the fact that more than a quarter began using drugs only after becoming homeless points to the role that the streets themselves play in worsening health outcomes.
The findings call for a fundamental shift in how society views homelessness and substance use. Kushel hopes policymakers, media, and the public will move away from “us vs. them” narratives that reduce unhoused people to caricatures.
“Homelessness is a housing problem,” Kushel concluded. “Drug use complicates it, but it is not the cause. Until we address housing, we are going to keep seeing people spiral on the street—whether they use drugs or not.”
As the debate continues, this study offers a critical data point: the majority of California’s homeless population are not regular drug users. The real challenge—and solution—lies not in blaming addiction, but in confronting the systemic failures that leave thousands of people without homes, treatment, or hope.
Source: UCSF Benioff Homelessness and Housing Initiative, Illicit Substance Use and Treatment Access Among Adults Experiencing Homelessness, JAMA, 2025
“just 37% of unhoused Californians reported regular use of illicit drugs”
Just? 37% is a lot in my book. And that’s if we can even believe the study that relied on responses from the homeless.
It’s definitely higher than the general population as noted in the article. But it also means that two-thirds aren’t – Seems like if we simply focus on drugs, we miss a huge part of the problem.
Moreover: “New drug use began for 27% after becoming homeless, suggesting homelessness itself is a driving factor for some.”
“Homelessness is a housing problem,” Kushel concluded. “Drug use complicates it, but it is not the cause. Until we address housing, we are going to keep seeing people spiral on the street—whether they use drugs or not.”
What if I said only 37% of the J6 protesters were violent?
But it also means that two-thirds weren’t.
It’s all in how it’s framed according to one’s biases.
It’s not a good analogy. One violent protestor can cause a lot of problems. If there’s only one homeless person using drugs, treating drugs isn’t going to address the bulk of the homeless problem.
Only 37% of our taxes were wasted by the Biden administration.
But two-thirds were spent wisely.
The take away point from the study:
“Homelessness is a housing problem,” Kushel concluded. “Drug use complicates it, but it is not the cause. Until we address housing, we are going to keep seeing people spiral on the street—whether they use drugs or not.”
From article: “A new study from the UCSF Benioff Homelessness and Housing Initiative (BHHI) . . . ”
So, that alone tells you the bias they would already have regarding the “cause”.
From article: ” . . . just 37% of unhoused Californians reported regular use of illicit drugs”
What people “report” is irrelevant regarding any valid conclusion, since it does not include those who don’t respond at all, weren’t randomly selected in the first place, give false answers, are too mentally-ill, high, or aggressive to even understand or respond to the question, etc.
Self-reported surveys are not valid regarding overall conclusions. That’s not how statistics work.
Also, how many abuse “licit” drugs (alcohol)?
Got it — here’s a balanced response you could use that pushes back while acknowledging nuance:
I hear the concern, but I disagree that self-reported surveys are inherently invalid for drawing conclusions — especially when done carefully. In this case, the researchers used two methods: a large-scale written survey and in-person, in-depth interviews, which helps validate the findings and minimize some of the common biases.
Yes, there are always limitations to self-reported data, but dismissing it entirely ignores how much of social science, public health, and criminal justice research relies on this methodology — because it’s often the only way to understand lived experiences.
“how much of social science, public health, and criminal justice relies on this methodology . . .”
(Now THERE’s a concerning acknowledgement. The problem is that these numbers are often reported to the public without examining how wildly-inaccurate they may be.)
Or on the converse, assuming it to be completely inaccurate out of hand
“4 out of 5 dentists surveyed recommend sugarless gum”
And here’s what I entertain – at least for my own amusement:
“And the other 95 dentists that we asked either didn’t respond to the survey, or said “don’t chew gum”.
Also, what, exactly, does that “one” dentist know that the other four don’t?” I won’t ask if any of them hold stock in a gum manufacturing company.
You’re comparing an advertising campaign to peer reviewed research?
Just pointing out how survey results can be reported.
Peer-reviewed should mean that they’ve already acknowledged the limitations. For all we know, the “gum survey” could pass a peer review as well.
And again, the organization which conducted this survey has a “mission” right in its title.
That’s not reporting and it’s not a survey
It is technically a survey, but my comment was intended to be humorous (and thought-provoking). Apparently, you didn’t experience either of those results.
But again, my point in my original 9:51 a.m. comment remains the same, and is unchallenged by you. Do you need me to repeat it?
“Only” 37% !
Like that is NOT a small number. This isn’t a presidential race and this isn’t the popular vote. You don’t “win” by not achieving 51% and then claim victory.
But you, DG, have reported here many times that in order to keep the ‘unhoused’ (cough, cough) numbers high, all sorts of categories of persons are counted who don’t live on the streets. I’ll bet my bippy — heck, both my bippies — that the drug abuse percentages are vastly higher amongst street people or the ‘visible homeless’, who are the persons who those of us concerned about public safety are concerned about the effects from.
There are all sorts of weasel words here — such as the fact that the wording is that the drug use is self-reported. There’s a stigma and legal issues around drug use, not to mention paranoia, so some may not report. Also, since some (read: the sane) ‘homeless’ programs require you to be sober, that might be yet another incentive to lie to the poll takers. Duh.
‘Homeless’ service providers have incentive to perpetuate false numbers in order to keep their own government-subsidized gig going. These people profit on failed concepts such as ‘housing first’ and ‘harm reduction’ which perpetuate the problems and cause ever-increasing subsidies that feed the Homless Industrial Complex. I consider these perpetrators to be as despicable of human beings as the drug dealers themselves.
“Where drug use does persist, it is often a survival mechanism . . . “People are using it as a coping tool—to stay safe, to stay awake, to make it all go away,” Yeah, that’s kind of what drug use does, while it kills you. So, now making excuses for meth use. Yeah, you gotta stay up, forever. No mention that the drug itself is perpetuating one’s own destruction.
“Homelessness is a housing problem” No it’s not, most especially not in regard to visible street people. If you believe that after building more brutalist architecture flats in Davis that there’s going to be a plethora of so-called ‘affordable’ units available that will be occupied by drug users off the streets and the ‘homeless’ numbers will dwindle, then I’ve got some beachfront property along a drainage ditch adjacent to the north end of F Street that I’d like to sell you.
I am not saying that we should shut down all homeless services. They should be available for those who do make the decision to stop. There are miracles.
“As the debate continues”,
You mean, as the delusion continues . . .
“ . . . the majority of California’s homeless population are not regular drug users.”
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA!!!!!!!!!!!!!!!
In space no one can hear you scream. But in Davis, people can hear me screaming.
“The real challenge—and solution—lies not in blaming addiction, but in confronting the systemic failures that leave thousands of people without homes, treatment, or hope.”
In other words, ‘$25 billion wasn’t enough. Give us more.’
This could be the worst article you’ve ever written, David Greenwald, and that’s saying something. But go ahead, counter with your ‘evidence-based’ evidence. The Homeless Industrial Complex has reason to crank out studies that justify keeping their funding going, and distribute that evidence-based evidence via it’s media tools.
I debated whether to respond – here goes…
Your rant reads like ideology, not analysis. The UCSF study is rigorous, peer-reviewed research — not propaganda from some imaginary “Homeless Industrial Complex.” You’re free to dismiss it, but that says more about your bias than the data.
As for “only” 37% reporting drug use — that number debunks the entire talking point that all unhoused people are addicts. And yes, it’s self-reported — just like every major survey in criminal justice, health care, and social science. If you want to throw out self-reported data, you might as well toss the Census too.
The real problem is simple: Housing costs have skyrocketed, wages haven’t kept up, and mental health services were gutted years ago. That’s why people are on the street — not because they’re all meth addicts who refuse help.
And no — more shelters with sobriety tests won’t fix it. Every city that’s reduced homelessness has done it the same way: housing first, services second. Everything else is just noise — or political theater for people more interested in complaining than solving anything.
You want to talk solutions? Start there. Otherwise, it’s just more yelling into the void.
Solutions? I am concerned about the effects of mass street people on the rest of society. Others can work on solutions for helping those trying to get out of the cycle of addiction. I am not opposed to that. What I have issue with is when mass money goes to ‘solutions’ that perpetuate the problem — which I see as the effects of street addiction on society.
You think I have an ideology bias, I think you have an ideology bias, but you think because of this study you are grounded in evidenced based evidence therefore your bias is confirmed as fact. As Journey sang: “Don’t Stop Believin'”. And I’m sure you won’t.
Another way to put it, almost 4 out of 10 homeless are chronic drug users.
Like I said, that’s a lot.
3.7 out of 10.0 Keith. This is science.