TRUSTED CONTENTEvery treatment website lists therapies. Almost none of them tell you which ones have strong evidence behind them, which are promising but unproven, and which are pleasant additions with no measurable effect on substance use outcomes. That distinction is the single most useful thing you can learn about this subject, and it is what this guide is built around.
Being honest about the evidence is not the same as being dismissive. Several approaches with thin outcome evidence are still worth having, for reasons this guide explains. But a family choosing a program deserves to know the difference, and so do you.
Who this is for, and how to read it
New to behavioral health. Read straight through and skip every green dashed box. The main text assumes no prior knowledge and no background in research methods.
Already working in treatment. The green dashed boxes cover what has moved recently, including a federal change that made one of the most effective interventions in the field practically deliverable for the first time.
A caution for everyone: nothing here tells you what any individual should have. Matching a person to an approach is clinical work. The purpose of this guide is to let you read a program description accurately and describe options without overselling them.
Progress saves in this browser only. Nothing is submitted anywhere and no sign-up is required.
The evidence scale used throughout
Every approach in this guide carries one of four ratings. They describe the weight of evidence that the approach improves substance use outcomes, not how much people enjoy it and not how good it feels to deliver.
Consistent benefit across multiple well-conducted studies. Recommended in clinical guidelines. A program offering none of these is worth questioning.
Real benefit shown, but in fewer studies, smaller effects, narrower populations, or mainly as an addition to other treatment rather than on its own.
Early trials are encouraging but the studies are small, short, or not yet replicated. Reasonable to offer alongside established care. Not reasonable to charge a premium for.
Either not studied properly, or studied and not shown to change outcomes. This does not mean worthless. It means the claim should be about enjoyment, engagement or wellbeing, not about treating the disorder.
One important caveat. A strong rating describes averages across populations. It does not mean the approach will work for a particular person, and an emerging rating does not mean it will fail them. Evidence guides the starting point; it does not replace clinical judgment.
The lessons
What “evidence-based” actually means, and what it hides
The phrase appears on nearly every treatment website, frequently attached to things that are not evidence-based at all. Understanding what it claims lets you read those pages properly.
An evidence-based approach is one tested in controlled research, usually randomized trials, and shown to produce better outcomes than no treatment or than an alternative. The strength of that claim depends on how many trials, how well conducted, how large the effect, and whether independent teams reproduced it.
Three things the label does not tell you
Whether it is delivered properly. Most researched therapies are manualized, meaning they follow a structured protocol, and their results come from trained clinicians delivering them with fidelity. A program can advertise a therapy and deliver something loosely resembling it. Asking how staff are trained and supervised in an approach tells you more than the name on the brochure.
How large the effect is. Statistically significant does not mean substantial. Some approaches produce modest average improvements that matter at population scale but feel small to an individual.
Whether it beats the alternatives. Many established therapies perform similarly to each other when compared directly. What consistently matters is that someone receives a structured, competently delivered treatment and stays in it, more than which of several good options it was.
The part that is easy to overlook
Across psychotherapy research, factors common to all approaches account for a substantial share of outcomes: the quality of the relationship between the person and the clinician, agreement on goals, and whether the person believes the treatment can help. This is not an argument that the specific approach is irrelevant. It is an argument that a person who trusts their counselor and is engaged in a good-enough treatment is likely to do better than a person receiving a technically superior treatment they dislike.
In practice
When a program lists twenty therapies, that is a marketing decision rather than a clinical one. Ask which two or three are the backbone of the program, who delivers them, and how those staff are trained. A program built on three approaches done properly beats one advertising twenty done casually.
Knowledge check: a program lists twenty therapies. What is the most useful follow-up question?
- Which is the newest?
- Which is most popular with clients?
- How many are holistic?
- Which two or three form the backbone, and how are staff trained in them?
Answer: which form the backbone, and how staff are trained. Fidelity of delivery matters more than the length of the list.
The strongest evidence in the field is not a talking therapy
If you rank everything in addiction treatment by weight of evidence, medication for opioid use disorder comes first, and it is not close. Methadone and buprenorphine substantially reduce the risk of death. No psychotherapy has demonstrated anything comparable.
This matters for how you talk about programs. A facility offering an impressive array of therapies but declining to support medication for opioid use disorder is omitting the intervention with the best evidence of keeping people alive. That is a legitimate thing for a family to ask about directly.
Opioid use disorder
- Methadone and buprenorphine are the two with mortality evidence behind them. Both work by occupying opioid receptors, reducing craving and withdrawal without the cycle of intoxication.
- Extended-release naltrexone blocks opioid effects rather than replacing them. It requires a period of abstinence before starting, which makes it harder to initiate, and the evidence is weaker than for the other two, though real.
Alcohol use disorder
Three medications are approved: naltrexone, acamprosate, and disulfiram. Naltrexone and acamprosate both have reasonable evidence for reducing drinking, with different mechanisms and different best-fit patients. Disulfiram works by making drinking unpleasant, so it depends heavily on supervision and motivation. Alcohol medications remain substantially underused relative to their evidence.
Stimulant use disorder
No medication is approved in the United States for cocaine or methamphetamine use disorder. This gap is why the next lesson matters so much: for stimulants, the strongest available intervention is behavioral.
Language note
Medication is not an alternative to therapy and the two are not in competition. The usual formulation is medication plus psychosocial treatment. Avoid framing that treats medication as a lesser or last-resort option, and never describe it as substitution.
Knowledge check: which intervention has the strongest evidence for reducing death in opioid use disorder?
- Residential treatment of 90 days or more
- Methadone or buprenorphine
- Cognitive behavioral therapy
- Twelve-step facilitation
Answer: methadone or buprenorphine. Both substantially reduce mortality, and no psychotherapy has shown a comparable effect.
Contingency management: the best-evidenced behavioral treatment almost nobody offers
Contingency management gives people tangible rewards, usually vouchers or gift cards, for verified behaviors: a negative drug test, attending sessions, taking medication as prescribed. It is unglamorous, it sounds almost too simple, and it has among the strongest evidence of any behavioral intervention in addiction, particularly for stimulant use disorder where no medication exists.
Why it works
Stimulants produce a powerful, immediate reward. Most treatment offers benefits that are delayed and abstract. Contingency management puts an immediate, concrete reward on the other side of the ledger, competing on the same timescale. Effects are strongest while the programme runs and diminish after it ends, which is a genuine limitation and an argument for longer duration rather than against the approach.
Why you rarely see it
Three reasons, and none of them are that it does not work:
- Funding rules. Federal fraud and abuse concerns long capped what publicly funded programmes could give, at levels well below what research showed to be effective.
- Reimbursement. Most payers have no billing route for it, though several state Medicaid programmes have obtained approval to cover it.
- Moral objection. A persistent belief that people should not be paid not to use drugs. This is a values argument, not an evidence argument, and it has cost lives during a stimulant overdose crisis.
Refresher — changed in January 2025
SAMHSA raised the annual incentive limit from $75 to $750 per patient per year for grant programmes that permit contingency management. The old $75 ceiling was far below the levels research associates with effect, which meant federally funded programmes were technically allowed to deliver the intervention while being unable to deliver it at a dose that works.
Incentives must be items, vouchers or gift cards rather than cash, and the guidance sets out safeguards around who assesses the target behaviour. If your understanding of contingency management is that it is capped at a level that makes it pointless, that changed.
Knowledge check: why does contingency management matter especially for stimulant use disorder?
- There is no approved medication for it, so the strongest option is behavioral
- Stimulant withdrawal is medically dangerous
- It is the cheapest intervention available
- It replaces the need for counseling
Answer: no approved medication exists. For cocaine and methamphetamine use disorder, contingency management is the intervention with the strongest evidence.
Cognitive behavioral approaches
The family of structured, skills-focused therapies that make up the backbone of most treatment programmes.
Identifies the thoughts, situations and feelings that lead to substance use and builds specific skills to handle them: recognising triggers, refusing offers, managing craving, planning around high-risk situations. Time-limited, structured, and among the most studied psychotherapies in existence. Effects are real and moderate rather than dramatic, and they hold up reasonably well after treatment ends.
A CBT variant focused specifically on identifying warning signs, high-risk situations and coping responses, and on treating a return to use as a learning event rather than a catastrophe. Frequently what people are describing when they say a programme is CBT-based.
Developed for borderline personality disorder and self-harm, where its evidence is strongest. Teaches distress tolerance, emotion regulation, mindfulness and interpersonal skills. A substance use adaptation exists and the evidence for it is moderate rather than strong. Most valuable where emotional dysregulation, self-harm or a co-occurring personality disorder is prominent, and it is frequently advertised well beyond that population.
Rather than disputing difficult thoughts and cravings, ACT teaches people to notice them without acting on them, while moving toward what they value. Evidence in substance use is promising and less extensive than for CBT.
A word about DBT specifically
Full DBT is a demanding programme: individual therapy, a skills group, telephone coaching between sessions, and a consultation team for the therapists. Many programmes advertising DBT run a skills group and nothing else. That may still be useful, but it is not DBT, and it is reasonable to ask which components a programme actually delivers.
Knowledge check: for which population is DBT’s evidence strongest?
- Everyone with a substance use disorder
- People with opioid use disorder specifically
- People with borderline personality disorder and self-harm
- Adolescents only
Answer: borderline personality disorder and self-harm. That is where it was developed and where the evidence is strongest. The substance use adaptation has moderate evidence.
Motivational approaches
Motivational interviewing is a way of having a conversation rather than a course of treatment. It works from the observation that ambivalence about change is normal, and that arguing someone out of it reliably produces the opposite of what you intend. Instead the clinician asks open questions, reflects back what they hear, and helps the person articulate their own reasons for change.
The evidence is good, particularly for engaging people early and getting them into treatment. Effects on longer-term outcomes are more modest, and it is usually a way in rather than the whole intervention. Motivational enhancement therapy is a structured, brief version, typically a few sessions incorporating personalized feedback.
Why this one matters even for non-clinical staff
You are not doing therapy on a call. But the underlying stance travels, and it is the difference between a call that ends in an appointment and one that ends in a hang-up. Arguing with someone about how serious their problem is, warning them where they will end up, or pressing for a commitment they are not ready to make tends to entrench the position you are trying to shift. Asking what concerns them, what they have already tried, and what they want does not.
The confrontational tradition
Older approaches built on confrontation, breaking down denial, and aggressive intervention were once common in American treatment and are still occasionally marketed. The evidence does not support them, and there is reason to think they increase drop-out. If a programme describes its approach as tough love or breaking someone down, that is a warning sign rather than a selling point.
In practice
The most useful habit borrowed from motivational interviewing is asking permission before giving information. “Would it help if I explained how the levels of care work?” gets heard. The same explanation delivered unasked frequently does not.
Knowledge check: what does the evidence suggest about confrontational approaches?
- They work best for severe cases
- They are equivalent to motivational approaches
- They are effective only in residential settings
- They are not supported, and may increase drop-out
Answer: not supported. A programme marketing itself on tough love or breaking down denial is describing something the evidence does not back.
Trauma-focused therapies, and what trauma-informed care is not
Trauma and substance use overlap heavily, and post-traumatic stress disorder is one of the most common co-occurring conditions in treatment populations. Two different things get called trauma work, and conflating them causes real confusion.
The two best-established treatments for PTSD. Prolonged exposure works by approaching trauma memories and avoided situations gradually rather than avoiding them; cognitive processing therapy works on the beliefs the trauma produced. Both are structured, time-limited, and supported by extensive trial evidence.
Eye movement desensitization and reprocessing has good evidence for PTSD and is recommended in major guidelines. The distinctive element, bilateral eye movements, remains debated: research has struggled to show that this component adds to the exposure and processing that EMDR shares with other trauma therapies. The treatment works. Which part does the work is less settled than its promotion sometimes implies.
Approaches designed to address both together, rather than requiring someone to be abstinent before trauma work can begin. Evidence supports treating both concurrently, and the older insistence on a long abstinence period before touching trauma has not held up well.
Trauma-informed care is not a therapy
Trauma-informed care is an organizational approach: recognizing that many people in treatment have trauma histories and running services so as not to re-traumatize them. Predictable routines, explaining what will happen, avoiding unnecessary restraint or humiliation, giving people choices where possible. It is good practice and it is not treatment for PTSD. A programme describing itself as trauma-informed may offer no trauma therapy at all, and it is worth asking which is meant.
Knowledge check: what is trauma-informed care?
- A specific therapy for PTSD
- An organizational approach to delivering services without re-traumatizing people
- Another name for EMDR
- A requirement for accreditation
Answer: an organizational approach. A trauma-informed programme may offer no trauma-focused therapy, so ask which is being described.
Family approaches and mutual aid
Two areas where the evidence is better than many people assume, and where families frequently arrive with the wrong expectations.
Works with the family member of someone who is not yet in treatment, teaching communication and reinforcement strategies rather than confrontation. It achieves markedly better rates of getting the person into treatment than confrontational intervention approaches, and it improves the family member’s own wellbeing whether or not their relative engages. This is the honest answer to “my son refuses to get help, what do I do,” and it deserves to be far better known than it is.
Works with the person and their partner together, on both substance use and relationship functioning. Good evidence where a supportive partner is available and the relationship is not violent.
For young people, family-based approaches such as multidimensional family therapy and functional family therapy are among the best-supported options, generally outperforming individual therapy alone. Adolescent treatment that does not involve the family is working with one hand tied.
A structured clinical approach designed to connect people with twelve-step fellowships, distinct from the fellowships themselves. Evidence has strengthened considerably: systematic review evidence indicates it performs at least as well as other established treatments and does better on sustained abstinence. Note carefully what this supports. It is evidence for the clinical intervention that links people to mutual aid, delivered as a manualized therapy, not a finding that any particular fellowship is required.
On mutual aid generally
Twelve-step fellowships, SMART Recovery, Refuge Recovery, LifeRing and others are free, widely available, and helpful to a great many people. They are peer support, not clinical treatment, and they work best alongside treatment rather than instead of it. People differ enormously in which one fits, and someone who bounced off one is not someone who has failed at recovery.
Knowledge check: a parent asks what to do about an adult child who refuses treatment. Which approach has the best evidence?
- CRAFT, which works with the family member directly
- A surprise confrontational intervention
- Waiting until the person reaches rock bottom
- Cutting off contact entirely
Answer: CRAFT. It gets more people into treatment than confrontational approaches and improves the family member’s wellbeing regardless of the outcome.
Holistic, complementary and emerging approaches
This is where honesty is hardest and most valuable. These approaches are heavily marketed, often beautifully photographed, and sometimes priced as though the evidence supported them. Below is what the research actually shows.
Read the ratings carefully. A low rating is a statement about measured effects on substance use, not a verdict on whether something is worth doing. Several of these improve how people experience treatment, and someone who stays in treatment because they enjoy part of it is getting a real benefit, just not the one being advertised.
Structured programmes such as mindfulness-based relapse prevention teach people to observe craving without acting on it. The best-evidenced of the approaches in this section, with reasonable trial support as an addition to standard treatment. Note the distinction between a structured, manualized programme and a facility simply offering a meditation session.
Reasonable evidence for improving mood, sleep and physical health during treatment, and some evidence for reducing substance use as an adjunct. Cheap, low-risk, and beneficial regardless. Worth offering; not worth billing as a primary treatment.
Small studies suggest benefits for anxiety, mood and possibly craving. Studies are generally small and short, and it is difficult to separate the effects of yoga from those of exercise and group activity. Pleasant, safe, plausible, and not yet established.
Delivered by credentialed therapists these are established professions with real clinical training. Evidence for effects on substance use outcomes specifically is limited, while evidence for engagement, emotional expression and treatment retention is more encouraging. Particularly useful for people who struggle to talk about feelings directly.
Widely offered in addiction settings. Trial evidence is mixed and generally weak, and better-controlled studies have tended to find little difference from sham acupuncture. Low risk, and some people find it calming, but it should not be presented as a treatment for the disorder.
Popular, photogenic, and frequently expensive. The research base is small and methodologically weak, and it does not currently support claims of improved substance use outcomes. People often value these experiences highly, which is worth something, but the marketing routinely outruns the evidence by a wide margin.
No reliable evidence of an effect on substance use. Massage is enjoyable and relaxing; that is a fair claim to make for it. Be alert to nutritional supplement protocols marketed as repairing brain chemistry, which sound scientific and are not supported.
Refresher — psychedelic and neuromodulation treatments
No classic psychedelic is currently approved by the FDA for any indication. An application for MDMA-assisted therapy for PTSD was declined in August 2024, with the agency requiring further trial work. Psilocybin programmes for depression are further along in the regulatory process and a decision may come in the near term, but that is for depression rather than addiction, and the position keeps moving. Verify current status before writing anything about it.
Evidence for psychedelics in addiction specifically remains at an early stage: small trials, encouraging signals, no approved treatment. Clinics offering these outside a trial are operating ahead of the evidence and, in many cases, ahead of the law.
Ketamine is a separate case. An esketamine nasal spray is approved for treatment-resistant depression, not for substance use disorder, and generic ketamine is widely used off-label. Its use in addiction is investigational, and ketamine itself carries misuse potential, which warrants particular caution in this population.
Knowledge check: which of these has the best evidence for substance use outcomes?
- Equine-assisted therapy
- Ear acupuncture
- Structured mindfulness-based relapse prevention
- Sound baths
Answer: mindfulness-based relapse prevention. The best-evidenced approach in this group, and note it means a structured programme rather than an occasional meditation session.
Review
Six questions spanning everything above. The first four check the main text. The last two are marked as refresher questions.
1. A programme treats mainly methamphetamine use. Which intervention should you expect to see?
- An approved anti-craving medication
- Contingency management
- Extended-release naltrexone
- Ear acupuncture
Answer: contingency management. No medication is approved for stimulant use disorder, so the best-evidenced option is behavioral.
2. What does an “insufficient evidence” rating mean in this guide?
- The approach is dangerous
- The approach has been banned
- The approach is never worth offering
- There is no reliable evidence it changes substance use outcomes
Answer: no reliable evidence of effect on substance use. Something can be enjoyable and worth offering while not being a treatment for the disorder.
3. A facility says it is trauma-informed. What have you learned about its trauma treatment?
- Nothing yet, since that describes how services are run rather than a therapy
- That it offers EMDR
- That it treats PTSD before substance use
- That its staff are trained in prolonged exposure
Answer: nothing yet. Trauma-informed care is an organizational approach. Ask separately what trauma-focused therapies are actually delivered.
4. Which factor consistently accounts for a substantial share of psychotherapy outcomes?
- The length of the programme
- Whether the setting is residential
- The quality of the relationship between person and clinician
- The number of therapies offered
Answer: the therapeutic relationship. Alongside agreement on goals and the person’s belief that treatment can help.
5. Refresher question: what changed for contingency management in January 2025?
- It was approved by the FDA
- The federal incentive limit rose from $75 to $750 per patient per year
- It became a required benefit under Medicare
- Cash payments were permitted
Answer: the incentive limit rose tenfold. The old ceiling sat well below the level research associates with effect. Incentives still cannot be cash.
6. Refresher question: what is the current status of psychedelic treatments for addiction?
- MDMA-assisted therapy is approved for substance use disorder
- Psilocybin is approved for alcohol use disorder
- Ketamine is approved for substance use disorder
- None are approved, and evidence in addiction remains early stage
Answer: none are approved. An MDMA application for PTSD was declined in 2024, psilocybin programmes target depression, and the position keeps changing, so verify before writing.
Job aid: questions that test a therapy claim
Use when reviewing a facility’s programme description, or when a caller asks whether somewhere is any good. None of these require clinical training to ask.
A quoted success rate with no definition and no follow-up period behind it is a marketing figure. Ask what counted as success, how long after discharge it was measured, and how many people they managed to reach.
Glossary
Select a term to reveal the definition.
ACT
Acceptance and commitment therapy. Teaches people to notice difficult thoughts and cravings without acting on them, while moving toward what they value.
CBT
Cognitive behavioral therapy. Structured, skills-focused work on the thoughts and situations that lead to substance use.
Contingency management
Providing tangible rewards for verified behaviors such as negative drug tests or session attendance. Strongest behavioral evidence in stimulant use disorder.
CRAFT
Community reinforcement and family training. Works with a family member to help engage a relative who is not yet in treatment, without confrontation.
DBT
Dialectical behavior therapy. Full delivery includes individual therapy, a skills group, between-session coaching and a therapist consultation team.
EMDR
Eye movement desensitization and reprocessing. An established PTSD treatment; the contribution of the eye movement component specifically remains debated.
Fidelity
How closely a therapy as delivered matches the protocol that was tested in research. Low fidelity is a common reason a named therapy does not produce researched results.
Manualized
Following a written, structured protocol. Most researched therapies are manualized, which is what makes them testable and reproducible.
Motivational interviewing
A conversational method that helps a person articulate their own reasons for change rather than arguing them into it.
Mutual aid
Peer-led support such as twelve-step fellowships, SMART Recovery, Refuge Recovery or LifeRing. Complementary to clinical treatment rather than a substitute for it.
Psychosocial treatment
The umbrella term for therapies and social interventions, as distinct from medication. Usually delivered alongside medication rather than instead of it.
Relapse prevention
A CBT variant focused on identifying high-risk situations and warning signs and planning coping responses.
Therapeutic alliance
The working relationship between person and clinician, including agreement on goals. Consistently associated with outcomes across all therapy types.
Trauma-informed care
An organizational approach to delivering services without re-traumatizing people. Not a therapy for PTSD.
Twelve-step facilitation
A structured clinical therapy designed to connect people with twelve-step fellowships. Distinct from the fellowships themselves.
References and further reading
Substance Abuse and Mental Health Services Administration. Using SAMHSA Funds to Implement Evidence-Based Contingency Management Services. Advisory PEP24-06-001, January 2025.
National Institute on Drug Abuse. Principles of Drug Addiction Treatment: A Research-Based Guide. National Institutes of Health.
Substance Abuse and Mental Health Services Administration. TIP 63: Medications for Opioid Use Disorder. HHS Publication.
Substance Abuse and Mental Health Services Administration. TIP 35: Enhancing Motivation for Change in Substance Use Disorder Treatment. HHS Publication.
Kelly JF, Humphreys K, Ferri M. Alcoholics Anonymous and other 12-step programs for alcohol use disorder. Cochrane Database of Systematic Reviews, 2020.
U.S. Department of Veterans Affairs and Department of Defense. Clinical Practice Guideline for the Management of Posttraumatic Stress Disorder and Acute Stress Disorder.
American Society of Addiction Medicine. The ASAM Criteria, Fourth Edition, Volume 1: Adults. 2023.
SAMHSA National Helpline: 1-800-662-HELP (4357), free, confidential, 24 hours a day. 988 Suicide and Crisis Lifeline: call or text 988.
This guide is educational and is written for people working in and around behavioral health services. It is not clinical training and not medical advice, and it cannot tell you which approach is right for any individual. Evidence ratings summarize the general weight of research at the time of writing and are not a substitute for current clinical guidelines or professional judgment. Treatment decisions should be made with a qualified clinician.
