Change Talk vs Sustain Talk in Motivational Interviewing

The fundamental insight of Motivational Interviewing is that clients argue themselves into — or out of — change. The strongest predictor of behaviour change isn’t how many reasons the clinician gives; it’s how much change talk the client themselves produces, and how strong that change talk is by the time the session ends.

This guide explains change talk and sustain talk, the DARN-CAT framework, how to evoke change talk deliberately, and how to respond to sustain talk without slipping into argument.

The Core Insight

Decades of MI research, starting with William Miller’s analyses in the 1990s, established that what clients say during a session predicts what they do afterwards. Sessions where the client produces a high ratio of change talk to sustain talk are followed by behaviour change at significantly higher rates than sessions where the ratio runs the other way (Magill et al., 2018, Journal of Consulting and Clinical Psychology).

This is why MI is structured around evoking. The clinician isn’t trying to convince — they’re trying to create the conditions where the client convinces themselves. The Spirit pillar of Empowerment is the philosophical commitment; change-talk evocation is the operational implementation.

The DARN-CAT Framework

Miller and Rollnick organise change talk into seven categories, grouped into preparatory and mobilising types:

Preparatory change talk — DARN:

Desire — “I want to…” / “I’d like to…”
Ability — “I could…” / “I think I’d be able to…”
Reasons — “It would be easier on my family if…”
Need — “I have to…” / “Something has to change.”

Mobilising change talk — CAT:

Commitment — “I will…” / “I’m going to…”
Activation — “I’m ready to…” / “I’m prepared to…”
Taking steps — “This week I called the clinic…” (the client describes action already taken)

The progression matters. A session where the client moves from Desire to Reasons to Need to Commitment is showing a recognisable trajectory toward change. A session that stays at Desire (“I’d like to drink less”) without ever reaching Commitment (“I will…”) is much less predictive.

Sustain Talk

Sustain talk is language that argues for the status quo. It’s the natural counterpart to change talk and almost always present alongside it — both sides of ambivalence speaking.

Examples:

“It’s the only thing that helps me relax.”
“Everyone in my family drinks like this.”
“I’ve tried before and it never works.”
“I don’t have time to exercise — when would I even fit it in?”

Crucially, sustain talk is not resistance. Resistance is a relational phenomenon (the client pulling away from the conversation or the clinician); sustain talk is a content phenomenon (one half of ambivalence, voiced). MI 3rd edition explicitly retired the term “resistance” partly to avoid this confusion.

When sustain talk dominates, it’s usually a signal that the clinician has been pushing — that the client is voicing the under-articulated half of their ambivalence in response to feeling unheard. The remedy is rarely more pressure.

How to Evoke Change Talk

Change talk doesn’t usually appear unprompted. The clinician’s job is to set the conditions and then notice and reinforce it when it appears.

Direct evoking questions

The most reliable lever. Ask questions whose only natural answer is change talk:

Desire: “What would you like to be different?”
Ability: “If you decided to do this, how might you go about it?”
Reasons: “What would be the upside of cutting back?”
Need: “What worries you about staying as you are?”
Commitment: “What’s something you’re willing to try this week?”

Note the framing: each question presupposes the change is on the table without insisting on it. A client can answer honestly without having to commit.

Looking back / looking forward

Looking back: “What was your life like before this got difficult?” — invites the client to access a pre-problem state, often surfacing values that point toward change.
Looking forward: “If we met in five years and things had gone really well, what would you be doing differently?” — bypasses current ambivalence by asking about a future state.

Ruler questions

“On a scale of 0 to 10, how important is making this change to you right now?”
“And why aren’t you at a lower number?” — this almost always elicits change talk. (Asking “why aren’t you higher?” elicits sustain talk; the framing matters.)

Selective reflection

When the client utters mixed change-and-sustain talk in the same sentence, reflect the change-talk side. The client typically continues developing it.

Client: “I’d like to cut down but it’s just so hard with everything going on at work.” Clinician: “You’d like to cut down.” (Reflecting only the change-talk fragment.) Client: “Yeah. I know I’d feel better. I’d sleep better. My wife would be happier…”

This is one of the highest-leverage MI moves. It’s also one of the easiest to miss; clinicians under time pressure tend to reflect the most recent or most emotional content rather than the change-relevant content.

Affirming change talk when it appears

Once the client has produced change talk, acknowledge it. Not with technique-y praise — with reflection that lets it land:

Client: “I’m tired of always feeling guilty about it.” Clinician: “The guilt is wearing on you, and there’s a part of you that’s ready for that to stop.”

Compare to a non-MI response: “That’s great that you want to change!” — which can feel patronising and shifts the locus from the client back to the clinician.

How to Respond to Sustain Talk

Arguing against sustain talk is the most common MI mistake. It triggers the discord effect: the client doubles down on the position they’ve voiced, even if they didn’t strongly hold it before. (“If you really wanted to change, you’d…” → “Maybe I don’t actually want to change.”)

The MI playbook is to defuse sustain talk, not contest it. Five techniques:

Simple reflection

Just reflect what was said. Sustain talk reflected often softens; sustain talk pushed against often hardens.

Client: “I’ve tried everything. Nothing works.” Clinician: “It feels like all the obvious options have run out.”

Double-sided reflection

Acknowledge both sides of the ambivalence, ending on the change-talk side (the second half of a double-sided reflection lands harder).

Clinician: “On one hand it’s been the easiest way to wind down at the end of the day. And on the other hand, it’s started costing you sleep and getting in the way of being present with your kids.”

Amplified reflection

Mildly overstate the sustain-talk content. The client often corrects in the change-talk direction.

Client: “There’s no way I can fit exercise into my schedule.” Clinician: “There’s literally not five minutes anywhere in the week.” Client: “Well, I mean, on Saturday mornings I sometimes have time…”

Use sparingly. If the client already feels unheard, amplification reads as sarcasm.

Reframing

Offer a different lens on the same content.

Client: “I’ve quit before and started again. I’ll just fail.” Clinician: “You’ve successfully quit before — and you’ve also learned something about what makes it harder to maintain.”

Coming alongside

When sustain talk is sustained, sometimes the strongest move is to stop pulling against it.

Clinician: “It might be that this just isn’t the right time for you. People know when they’re ready, and you might not be.”

This isn’t surrender. It’s an honest naming of where the client is, which paradoxically often opens space for them to move.

What to Track in Sessions

Practitioners working on this aspect of MI find it useful to track, per session:

Change-talk-to-sustain-talk ratio. Higher is better, but absolute counts matter too — a session with 2 change-talk utterances and 0 sustain talk doesn’t tell you much.
Strength of change talk. Mobilising (CAT) carries more predictive weight than preparatory (DARN). A session with five Desire statements and zero Commitment is preparatory; a session with two Commitments and one Taking-Steps is meaningfully further along.
Where in the session change talk appeared. Late-session change talk has more behavioural follow-through than early-session change talk that wasn’t built on.

Common Mistakes

Reflecting the most recent thing instead of the change-relevant thing. Selective reflection requires you to listen for the change-talk fragments in mixed utterances, which is a different skill from general reflective listening.
Asking evoking questions but not following the change talk. A clinician asks “what would the upside be?”, the client gives one sentence of change talk, and the clinician immediately moves to the next question. The change talk needs space to develop.
Treating change talk as a target to be hit rather than something to be cultivated. Asking ruler questions in rapid succession to “extract” change talk feels manipulative and usually triggers sustain talk in response.
Confusing sustain talk with resistance. Sustain talk is information about ambivalence; resistance is a signal about the relationship. They need different responses.

Frequently Asked

How much change talk is enough?

There’s no fixed threshold. Magill et al.’s 2018 meta-analysis found that the percentage of change talk in a session is a meaningful predictor of behaviour change, but the effect size varies by population and behaviour. A useful rule of thumb: by the end of an MI session, the client should have voiced multiple distinct change-talk statements, and at least one of them should be mobilising (CAT) rather than just preparatory (DARN).

What if my client gives no change talk at all?

Possible signals: (1) you haven’t asked evoking questions yet; (2) you’ve been arguing against sustain talk and the client has dug in; (3) the client genuinely isn’t yet considering this change, and the session needs to focus on engaging rather than evoking. The fix depends on which one is happening — a transcript review usually shows quickly.

Can sustain talk be useful?

Yes. Sustain talk surfaces the obstacles the client will actually face if they try to change. Suppressing sustain talk to keep the session feeling positive can leave the client with an unrealistic plan they then can’t execute. The goal isn’t zero sustain talk — it’s a session where the change-talk-to-sustain-talk ratio shifts in the change direction over time.

Is “Discord” the same as “Resistance”?

In MI 3rd edition Miller and Rollnick separated these concepts. Discord is dissonance in the relationship — the client is pulling away from you. Sustain talk is content about the behaviour — the client is voicing one side of ambivalence. They look similar from the outside but have different sources and need different responses. Discord often means the clinician has overstepped; sustain talk often means the client is processing.

Want to practise evoking change talk? The MI Practice Lab tags every change-talk and sustain-talk moment in the transcript, lets you filter the timeline by category, and surfaces the change-talk you missed reflecting. Start a free trial — 5 minutes, no card required.