A thought record is the workhorse of cognitive behavioral therapy. Seven fields, five to ten minutes: a moment that produced a strong feeling, the thought behind it, evidence on each side, a more accurate conclusion. The format was systematised inside Beck's cognitive therapy for depression in 19791 and turned into the self-help worksheet most people now recognise by Greenberger and Padesky's Mind Over Mood in 1995.2
The point isn't journaling. It's to interrupt the loop between a situation, the automatic thought it triggers, and the feeling that follows, at the one place the loop is open to inspection: the thought.
A 2010 meta-analysis of 46 trials found CBT with structured homework produced effect sizes roughly double CBT without it.4 The thought record is the most common form that homework takes.
A worked example
You sent a friend a voice message Tuesday morning asking to meet that weekend. By Thursday evening she hasn't replied. Your stomach tightens every time you check the app. We'll fill in the seven fields the way the Colors reframe flow asks for them.
1. Situation
Write what happened, without interpretation. The bar is what a security camera would record.
Sent Anna a voice message Tuesday morning asking about the weekend. It's now Thursday 9pm and she hasn't replied. She's been online today.
Notice what's not in there: no "she's ignoring me", no "she's busy". Both belong in the thought field. Beck Institute training materials5 are firm on this because conflating the two is the most common reason a thought record fails.
2. Automatic thought
The exact words the mind produced. Not a tidied-up version.
She's annoyed with me about the thing last week. She doesn't want to see me anymore.
If several thoughts came through, pick the one with the strongest charge. Greenberger and Padesky call this the hot thought.2 A thought record can only restructure one at a time.
3. Reaction
The emotion, its intensity from 0 to 100, and any body sensations. Colors asks for the emotion as a tag, the intensity as a slider, the body part as free text.
Anxious 75. Some sadness 40. Tight stomach, slight pressure behind the eyes.
You'll re-rate at the end and want to compare. Putting a number on a feeling is also part of the regulation: it moves the feeling from a vague cloud to something with edges.
4. Evidence for
What actually supports the automatic thought? Specific events, specific words, observable behavior. No interpretations.
Last week's dinner ended early and she seemed quiet on the way home. She replied within an hour to my last three messages and this one has been two days. She's been active on Instagram today.
The temptation is to pad the list with feelings ("I just sense something's off") or skip the field because the case feels obvious. Both fail the exercise. Write the strongest version you can; the next field is where it gets tested.
5. Evidence against
The same standard, in the other direction. What facts cut against the automatic thought?
She's mid-deadline this week; she mentioned it on Sunday. She's replied late before during busy weeks and it's never meant anything. The dinner ending early was the restaurant closing, not a fight. She used a heart emoji on my last story this morning.
This is usually the longest field and the one people resist most. The mind that produced the hot thought isn't eager to undermine it. Useful prompt: what would I tell a friend in this exact situation? The answer almost always lands here.
6. Balanced outcome
A single sentence that integrates both columns. Not a positive reframe; an accurate summary of what the evidence supports.
She's probably swamped with the deadline and will reply when she surfaces. There's a small chance something's off from last week, and if there is we'll talk about it when we meet.
Then re-rate the emotion. In a real reframe this often drops from 75 to 35, not to zero, because the situation isn't resolved. CBT calls this accurate thinking, not positive thinking. Aiming for zero anxiety produces brittle reframes that collapse the moment new ambiguity arrives.
7. Distortions
Tag any of the 14 cognitive patterns that drove the automatic thought. In our example: mind reading (assuming Anna is annoyed without checking), prediction (she doesn't want to see me anymore), and personalisation (reading her silence as caused by you).
Most automatic thoughts run two or three distortions stacked together; spotting the stack does more work than challenging the conclusion. The full set is in cognitive distortions.
What makes thought records work, and what makes them fail
A few things determine whether the exercise actually moves the feeling:
- Write it at the time, not hours later. Once the moment has passed, the emotion has already started to change and the thought gets reconstructed instead of caught raw.
- Use the exact words that appeared. "I felt like she didn't want to see me" is already a paraphrase. The actual thought was "She doesn't want to see me anymore." The distortion is in the wording.
- Do the evidence step before you argue. Jumping straight to "but that's not true" usually produces reassurance that doesn't last. The re-rating at the end only feels real when you've done the work.
- Re-rate the emotion at the end. Without the second number there is no feedback that the exercise did anything. If the intensity doesn't drop at all, the hot thought you wrote probably wasn't the real one.
Thought records aren't a replacement for therapy when the thoughts are tied to active depression, panic, OCD, or PTSD. They're a tool that works inside therapy and after it. For everyday automatic thoughts, the small repeated ones that drive most ordinary distress, they're enough.
How Colors handles it
Colors' reframe flow walks through these seven fields in order after you've logged a feeling. Situation and automatic thought are free text. The reaction carries over the emotion you already tagged, plus a 0–100 intensity slider and an optional body sensation. Evidence for and against are paired free-text fields. The balanced outcome prompts a re-rating, and distortion tags use the same 14 patterns picked from a multi-select. Over weeks the entries become a searchable history of which thoughts and which distortions recur, which is what makes the format more useful in an app than on a worksheet.
The work itself is unchanged from what Beck and his colleagues laid out in 1979.1 A situation, a thought, a feeling, the evidence on both sides, a conclusion that holds up. Five to ten minutes, done at the time, repeated.
Frequently asked questions
What is a thought record?
A thought record is a short written exercise from cognitive behavioral therapy where you write down a situation, the automatic thought it triggered, the emotional reaction, the evidence for and against the thought, and a more balanced conclusion. The format was developed inside Beck's cognitive therapy in the 1970s and popularised for self-help by Greenberger and Padesky's Mind Over Mood in 1995.
How long does a thought record take to fill in?
Five to ten minutes once you have the format memorised. The first few will be slower because you're learning the prompts. Therapists typically assign one or two a day between sessions, and the meta-analytic evidence suggests this kind of structured homework roughly doubles the effect size of CBT compared with sessions alone (Kazantzis et al., 2010).
Do I have to fill in every field?
Yes, in order. The fields are designed to slow the thinking down at points where it usually accelerates: between situation and reaction, and between reaction and conclusion. Skipping the evidence steps and jumping to a 'balanced outcome' tends to produce reassurance rather than restructuring, which doesn't generalise outside the moment.
What's the difference between a thought record and a journal entry?
A free-form journal records what happened. A thought record forces a specific structure — automatic thought, evidence both ways, conclusion — that maps onto how CBT thinks the loop between situation, thought, and feeling actually breaks. Colors uses the structured form for its reframe flow and leaves free text for the situation field, so you get both.
Can a thought record make things worse?
It can if you treat it as a place to argue with yourself or write down catastrophic predictions without then doing the evidence step. The format only works when you complete it. For trauma-related intrusive thoughts and for active depression severe enough to interfere with daily functioning, thought records are usually done with a clinician rather than alone.
Not medical advice
This article is for informational and educational purposes only. It does not constitute medical advice and should not replace consultation with a licensed mental health professional. If you are in crisis, please contact emergency services in your country immediately.
Crisis lines: US — 988 Suicide & Crisis Lifeline · UK / Ireland — Samaritans 116 123 · EU — Befrienders Worldwide
Last reviewed: May 2026.
References
- Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive Therapy of Depression. Guilford Press.
- Greenberger, D., & Padesky, C. A. (2016). Mind Over Mood: Change How You Feel by Changing the Way You Think (2nd ed.). Guilford Press.
- Beck, J. S. (2011). Cognitive Behavior Therapy: Basics and Beyond (2nd ed.). Guilford Press.
- Kazantzis, N., Whittington, C., & Dattilio, F. (2010). Meta-analysis of homework effects in cognitive and behavioral therapy: A replication and extension. Clinical Psychology: Science and Practice, 17(2), 144–156. doi:10.1111/j.1468-2850.2010.01204.x
- Beck Institute for Cognitive Behavior Therapy. Thought records and CBT worksheets. beckinstitute.org