Motivational Interviewing for Substance Use

Motivational Interviewing is a way of conducting a conversation so that a person voices their own reasons for changing a substance use behaviour, rather than hearing yours. It was first described in exactly this population, and it remains the most widely used method for the part of substance use work that happens before treatment starts: the appointment where someone is not yet sure they want to be there. The evidence supports it as a way into change. It does not support it as a superior treatment once someone is already engaged, and knowing the difference changes how you use it.

Does Motivational Interviewing work for substance use?

The honest answer has two halves, and most pages on this subject only publish the first one.

A Cochrane review published in 2023 pooled 93 randomised trials covering 22,776 people who used drugs, alcohol or both (Schwenker and colleagues, Cochrane Database of Systematic Reviews, 2023). Against no intervention, MI showed modest benefit at short-term follow-up. Against assessment and feedback, it showed a small benefit at medium and long-term follow-up. Against treatment as usual or another active treatment, the reviewers concluded that MI “may make little to no difference to substance use”. They were careful about the certainty of all of it, describing “moderate to no confidence in the evidence”.

That pattern is not new. Project MATCH, the largest psychotherapy trial ever run in the alcohol field, randomised 952 outpatients and 774 aftercare clients across three twelve-week individually delivered treatments, one of which was Motivational Enhancement Therapy (Journal of Studies on Alcohol, 1997). It found significant and sustained improvement in drinking outcomes across all three arms, with little difference between them. The matching hypothesis the trial was designed to test largely failed.

Read together, those two findings say something specific and useful. MI is not the ingredient that makes a treatment work better than another treatment. It is the thing that gets a person to the point where any treatment can work at all, and that is where its value sits: intake, brief contacts, the appointment nobody chose, the fifteen minutes on a ward, the conversation after a relapse. Use it as an engagement method and the evidence backs you. Use it as a claim that your service outperforms the one down the road and it does not.

Why the method came out of the alcohol field

MI was first set out in a 1983 paper by William Miller, “Motivational interviewing with problem drinkers”, in Behavioural Psychotherapy (volume 11, pages 147 to 172). The setting matters, because of what it was written against.

Addiction treatment at the time was substantially confrontational. Denial was treated as a feature of the condition, so breaking it down was treated as clinical work: the label was pressed, the consequences were listed, the client’s account was challenged. Miller’s observation was that therapist style predicted drinking outcomes, and that confrontation predicted them in the wrong direction. The more a counsellor argued, the more the client argued back, and the arguing back was reliably followed by drinking.

That history is still in the room. Many clients have met the confrontational version of this conversation already, from a previous service or from a family member, and they arrive with a script prepared for it. Noticing that they are answering someone else before they are answering you is often the whole of the first session.

What changes when the behaviour is substance use

The method is the same one described across the four processes and OARS. Four things about this population change how it plays out.

The righting reflex is at its strongest. The urge to put someone right is proportional to how visible the harm is, and in substance use the harm is often extremely visible: the liver results, the missed contact, the overdose two weeks ago. So the pull to warn peaks precisely where warning is least effective. Every clinician knows the theory. Very few resist it when someone in front of them is doing something that might kill them.

The client has heard the argument. Whatever you are about to say about their health, their children or their job, someone who loves them has already said it, probably more forcefully. Adding your version to the pile does not move anything. It just locates you among the people who talk at them.

Labels do work in the room, and rarely the work intended. “Alcoholic”, “addict”, “in denial” and “non-compliant” all invite the client to argue about the category rather than the behaviour. The category is a debate you can lose. The behaviour is not.

The ambivalence is real. Substances are doing a job: sleep, pain, social ease, the only reliable hour of the day. Treating ambivalence as an error of reasoning, something to be corrected with better information, misreads it. The client is not confused. They are holding two accurate things at once, which is the definition of the state MI was built to work with.

The appointment nobody chose

A large share of substance use work is mandated in some form: probation and community orders, drug courts, child protection plans, employer referrals, conditions of housing or of continued prescribing. In many services this is the normal case rather than the difficult exception.

The mistake is arguing about the mandate. It is not yours to remove, the client knows it, and defending it puts you on the wrong side of the table before anything else has happened. What can be negotiated is what happens inside the hour.

Client: “Social services sent me. You already know what you’re going to write. So write it, and I’ll go.”

Worker: “You’ve come in expecting this to be a form-filling exercise with a conclusion already decided.” (A reflection, not a denial. Contradicting the expectation confirms it.)

Client: “Isn’t it?”

Worker: “I do have to write something, and I won’t pretend otherwise. What goes in it isn’t decided. You know more about your drinking than anyone in this building does, and I’d rather the report was accurate than quick.” (Honesty about the constraint, then autonomy: the account is his.)

Client: “It’s not even about the drinking. It’s about the flat.”

Worker: “Tell me about the flat.”

The last line is the point. He has just offered a focus, and it is not the one on the referral. Following it is not a detour from the work; it is the only route to any of the rest of it. A worker who steers back to units per week loses the only thing the client volunteered.

Where sustain talk comes from in this work

Sustain talk climbs for a reason, and in substance use the reason is usually the previous sixty seconds rather than the client’s personality. Three reliable triggers:

Arguing for change. The client takes the other side, because that is what conversations do. This is the righting reflex arriving as a boomerang.
Premature planning. Detox dates, meeting timetables and referral forms produced before the person has decided anything. The plan belongs to you, so the objections do too.
The label. Any sentence that assigns a category invites a defence of the category.

How each of these sounds, and what to do when it appears, is set out in change talk and sustain talk. What matters here is the diagnostic use: rising sustain talk is feedback on the conversation, available in real time, and it is the most useful signal in the room.

Readiness is worth reading the same way. The stages of change came out of this field and describe who is sitting in front of you well, but placing someone in a stage is not a prerequisite for practising MI, and Miller and Rollnick have said so directly.

What goes wrong

The move	Why it backfires
Listing consequences	Supplies the argument for change, so the client supplies the argument against
Insisting on the label	Turns a conversation about drinking into a debate about a word
Assessment first	Twenty minutes of structured questions puts the client in the passive seat before any relationship exists
Abstinence as the only acceptable goal	Ends the conversation with anyone not ready to accept it, which is most people at first contact
Correcting the relapse account	A relapse the client feels judged for is a relapse they will not describe accurately next time

How to practise this

Reading about resisting the righting reflex does nothing for the moment it actually fires, which is when someone describes drinking that frightens you and you have half a second to decide what leaves your mouth. That is a reflex, and reflexes are retrained by repetition, not by understanding.

The MI Practice Lab runs voice conversations with AI clients written for this work, including mandated referrals and flat refusals of the premise. Afterwards your change talk and sustain talk come back marked up with the moment each appeared, so you can see which of your responses preceded the climb, alongside OARS-tagged playback and Critical Moments. Worked scripts for several substance use scenarios are also on the MI roleplay page if you would rather rehearse with a colleague.

Frequently Asked Questions

Is Motivational Interviewing effective for substance use?

It has good evidence as an engagement method and weaker evidence as a stand-alone treatment. The 2023 Cochrane review of 93 trials and 22,776 participants found modest benefit against no intervention and a small benefit against assessment and feedback, but concluded that MI may make little to no difference compared with treatment as usual or another active treatment. The practical reading is that MI earns its place at first contact, in brief interventions and at moments of disengagement, rather than as a claim that it outperforms other therapies.

How is MI used in addiction treatment?

Most often at the front of the pathway and at its friction points: assessment appointments, brief contacts in hospital or primary care, mandated reviews, and the conversation after a relapse. It is also used as a style layered over other treatment rather than as a separate block of sessions, which is how Motivational Enhancement Therapy was structured in Project MATCH.

What do you do when a client is only attending because they have to?

Do not argue for the mandate and do not pretend it is absent. Name the constraint honestly, make clear what is still the client’s to decide, and negotiate what happens inside the appointment rather than whether the appointment should exist. Whatever the client raises first, including something apparently unrelated to the substance, is usually the only available route into the rest of the conversation.

Should you use the word alcoholic or addict with a client?

Generally not, unless the client uses it first about themselves. Labels invite a debate about whether the category fits, which is a debate the clinician can lose and which produces nothing useful either way. Describing the behaviour and its effects keeps the conversation on ground the client cannot dispute.

Does MI work with people who are not ready to stop?

That is the population it was designed for. MI does not require the client to accept a goal of abstinence, or any goal, before the conversation can begin. Its purpose is to draw out and strengthen whatever reasons for change the person already holds, which is why it is most useful precisely when readiness is low.

How do you use MI after a relapse?

Treat the account as information rather than as a failure to be examined. A client who expects judgement will edit what they tell you, and an edited account is clinically useless. Reflecting what happened without evaluation, then asking what they made of it, tends to produce both a more accurate picture and more change talk than reviewing what went wrong.

Want to practise the mandated appointment before you are in one? The MI Practice Lab lets you hold voice conversations with AI clients written for substance use work, then shows you exactly where your sustain talk climbed and what you said just before it. Start a free trial: 5 minutes, no card required.