Miracle Question Script for This Hour

It’s 3:18. Rae has said “lighter” twice. You still do not have a sign you could point a camera at. The next line you offer is the miracle question, or you do not ask it.

Adapt every line to the person in front of you, your formulation, and your protocol. A composite exchange is a rehearsal, not a recording of a client, and it does not replace supervision or your crisis procedure.

What the miracle question is in counseling

In counseling, the miracle question is the sequence in the box, not a definition you recite and not a wish you assign. You get permission. You ask for the first ordinary difference in the morning. You stay with what a camera could catch. You write what they said.

The 2025 Solution Focused Brief Therapy Association treatment manual describes the sequence as an invitation to a detailed, client-generated description of what will be different when what they want is happening more often. It records that the question was first used at the Brief Family Therapy Center in the early 1980s, invented from an exchange between a client and Insoo Kim Berg. Berg had asked what would need to happen for the talking to be useful. The client said maybe only a miracle would help. Berg answered, in the wording the manual quotes from De Jong and Berg (2013, p. 91): “Okay, suppose a miracle happened, and the problem that brought you here is solved. What would be different about your life?” (SFBTA treatment manual, 3rd ed., 2025).

That is the job. A morning difference in their words. Not a theory of the problem, and not a goal you type while they are still talking.

The same manual says BFTC experience taught the team that clients who constructed clear, detailed miracle pictures were more motivated to change and made more progress over fewer sessions. Treat that as the manual’s account of clinic experience, not as a trial result you can quote to a payer. You still do the work of getting a picture that is concrete enough to write down.

A useful picture, in that manual’s account of practice since the 1980s, has three characteristics: the presence of desirable behaviors rather than the absence of problems; a beginning step rather than a final result; and a realistic description of what others will notice as well as what the client will notice. Rae’s “lighter” fails the first two. “Blinds open, phone face down” meets them. “My roommate would see the blinds up” meets the third, and the limit (“do not ask her”) belongs in the note with the sign.

SFBT miracle question wording you can learn

The wording the manual records from de Shazer (1988, p. 5) is this, asked deliberately:

“Now, I want to ask you a strange question. Suppose that while you are sleeping tonight and the entire house is quiet, a miracle happens. The miracle is that the problem which brought you here is solved. However, because you are sleeping, you don’t know that the miracle has happened. So, when you wake up tomorrow morning, what will be different that will tell you that a miracle has happened and the problem which brought you here is solved?”

Learn that shape so you know what the question is for. The lines in the session box are original session language for the same job. They are not a reprint of the 1988 sentence, and they are not a shorter version you invent in the moment because you are nervous.

de Shazer, in a 2002 question-and-answer note, said he emphasized the exact wording, including the pauses, because that wording was the product of almost 20 years of experiments and experience, and other versions had not gotten as consistently useful responses. He also said that knowing the wording freed him to watch and listen, because he could not think up the sentence and listen at the same time (de Shazer, July 2002).

That is the reason you rehearse the setup and the question before the hour, not during it. If you are still composing the sentence, you will miss the flinch, the pass, or the feeling word that needs a camera question.

Say the setup first. “You can pass on this. If you pass, we stay with this week.” A pass is an answer. Write it. Do not rephrase the question until they agree. Pressure dressed up as curiosity is still pressure.

Then the question, as written. Pause after “tomorrow morning.” Pause before you ask for the ordinary thing. The pause is part of the wording, not dead air you have to fill.

Miracle question examples

Use the Rae exchange as the rehearsal, then change the sign to whatever this client actually says. The structure stays. The content does not.

Feeling word, then a camera. They say “I’d be happy” or “lighter” or “calm.” You do not accept that as the answer. “Skip the feeling. What would a camera catch in the first ten minutes?” If the second answer is still a feeling, ask once more for an action, then stop. A third demand is an argument.

A sign that is already small. “The blinds would be open. The phone would still be face down.” That is enough. Do not upgrade it to a new job, a repaired marriage, or a week without the problem. The manual’s beginning-step test is the check: if the picture is the final result, ask what they would notice in the first ten minutes, not what the whole life would look like.

A person who would notice. “Who else would notice, and what would they see you doing?” You want an action, not a feeling the other person would have. If they cannot name a person, do not supply one. A coworker you invent is your hypothesis. If the named person should not see the work, say that before you write the line. A roommate who reads the chart over their shoulder is not a useful observer. You can still keep the camera sign.

A piece that has already happened. “Which piece of that picture has already shown up once in the last two weeks?” Thursday’s half glass of water is the piece. You write it. You do not open a separate search for every exception in the month. If they say nothing like it has happened, accept that and stay with tomorrow morning. Do not insist.

A pass. “I’d rather stay with this week.” You stay with this week. The note says they passed. You do not explain why the question would have helped.

Follow-ups after the first sign

The manual calls this question the opening of a co-constructive sequence. You connect to their words for what they want different, and you keep inviting detail. Clients will mix that detail with what they do not want. You punctuate the part that is a sign, and you leave the problem talk in the room without building the next question out of it.

Pick two follow-ups. The list in the box is the set. A third follow-up is how a short question becomes an interview they did not agree to.

If they say “I don’t know,” or they go quiet, the manual’s advice is to assume you asked a hard question and to remain silent. Most often the “I don’t know” is followed by a pause and then a beginning answer. Fill the pause and you take the turn back. Wait. If the silence is shutdown rather than thought, you already have a reason to stop. Name it: “We can leave this. We stay with this week.”

Move a vague answer toward something you could see. The manual’s practice note is to invite observable terms, because a feeling word does not tell you, them, or anyone else whether the sign happened. Relationship questions tend to do that work, because they ask what another person would notice. Use one. Do not interview that person from the chair.

Stop when you have a sign, a who-would-notice line or an explicit skip, and either a piece that already happened or a clear “nothing like that yet.” That is a picture. More questions will sand it down.

Anchor it with one scale question

After the picture, one number. The manual says it is common, after that sequence, to ask clients to rate where they are on a scale where 0 equals the problem at its worst and 10 equals the miracle they have just described. de Shazer and colleagues (2007) called that form the miracle scale. In the manual’s dialogue, the interviewer then asks what tells them they are at that number, and the answers are exceptions tied to the picture, not a lecture about effort.

The anchor in the session box is smaller on purpose. 10 is the first hour they just described. 0 is this morning. You are not asking them to score the worst day of their life in the same breath as a blinds-and-phone picture. If you use the manual’s wider poles, say so, and take both ends from their words, not from yours.

Ask where they are today. Write the number they give. Ask what holds it off the floor only if they have already named it, as Rae named the bus walk. Then stop.

Do not argue with the number. Do not ask for a jump of more than one point. Do not turn a 3 into a talk about trying harder. If they want to stay with what one point up would look like, that is a scaling conversation, and it lives with the other stems on the solution-focused therapy questions page. It is not a second technique you staple onto this question because the hour still has twenty minutes.

The number is not a symptom score. Do not drop it into a PHQ-9 field, a GAD-7 field, or an outcome graph. If your protocol requires a measure, administer the measure. The anchor does not replace it.

Write the note line

One sentence, in their words. The type of question. The first sign. Who would notice, and that you did not interview them. The picture was not assigned as a goal. The number, if you asked.

A line that says “used a future-difference question, client engaged, will continue” could sit on any chart. It will not help you next week, and it will not tell a colleague what was asked. Rae’s line will. You can see the blinds, the phone, the roommate limit, and the 3.

Do not paste the picture into the goal field. The picture was a question, not a contract. If you and they later want a goal, that is a different conversation, on a plan you build with them, not on a morning they were invited to imagine.

If they passed, the note is shorter. “Offered a future-difference question. Client passed. Stayed with this week.” That is a complete line. Do not fill the blank with a sign they did not give you.

When you want that sentence in the chart without retyping the sign, try Emosapien on the next session. You keep their wording.

When you skip the question

Skip it when a safety concern is in the room. Use the risk process you already use. A morning picture is not a screen, and a calm voice is not a substitute for the question your protocol requires. You can ask about a sign after you have the safety answer, not before it.

Skip it when they pass. The setup promised you would stay with this week. Keep the promise.

Skip it when the future picture lands as a demand to feel better. You will hear it in the second answer: shorter, flatter, or a performance of the life they think you want. Stop. Return to today. Say, “That landed as pressure. We can leave it.”

Skip it when hypotheticals confuse more than they help. Some clients cannot use a suppose-while-you-sleep frame, and pushing the frame is not clinical courage. Stay with what they are already doing that a worse week would have stopped, or move to a different stem. Coping, exception, and scaling questions are different tools. They are on the solution-focused therapy questions page. Do not teach them inside this question.

Skip the compliment and the between-session task unless the hour produced one sign worth noticing. Noticing is not the same as performing the better morning. If you assign anything, assign observation of the sign they already named, not a rehearsal of the whole picture.

A course will not put the wording in your mouth. If you want hours, a format, and a certificate name, that comparison is on the solution-focused therapy training page. Book it after you have asked this question once, not instead of asking it.

Download the Solution-Focused Question Cards

The session box is the sequence for this hour. The Solution-Focused Question Cards add a chooser page and five other cards, each with a one-line note stem. Print the miracle card if you want the setup and the follow-ups in your hand. Ask one question. Write their words. Copy the stem. Do not run the file as a script, and do not send it home as homework.

References

• de Shazer, S. (2002, July). Response on technique and question wording. Solution Focused Brief Therapy Association. https://www.sfbta.org/techniques_q-a.pdf
• Solution Focused Brief Therapy Association Research Committee. (2025). Solution Focused Therapy Treatment Manual for Working with Individuals (3rd ed.). Solution Focused Brief Therapy Association. https://www.sfbta.org/Solution%20Focused%20Therapy%20Treatment%20Manual%20for%20Working%20with%20Individuals%20-3rd%20Edition-%20final.pdf