Cognitive Behavioral Therapy works on one idea: it isn't events that flatten us, it's the thoughts we have about them — and thoughts can be caught, tested, and changed.
CBT tests whether a thought is actually true, then changes it — and the evidence for it is strong, with one honest caveat: effect sizes shrink once you control for publication bias.
Well-supported = backed by replicated randomized controlled trials · Promising = smaller studies or a single trial not yet replicated. Self-help, not a diagnostic or treatment tool.
Cognitive Behavioral Therapy was developed by Aaron T. Beck in the 1960s, building on Albert Ellis's earlier work. Its core claim: it's rarely the situation itself that produces distress, but the automatic thought about the situation — and because that thought is a mental event, not a fact, it can be caught, checked against evidence, and revised. Later additions, especially behavioural activation, extended the model with a second lever: deliberately changing what you do, not just what you think, since avoidance itself maintains low mood.
Reach for this when a specific thought is looping and you can genuinely test it against evidence — "everyone thinks I messed up," "I always fail at this." It's also the tool for depression's motivation problem: when waiting to feel like doing something has stopped working, behavioural activation is the deliberate alternative. It's less suited to a thought that's actually accurate about a genuinely bad situation — more on that below — and less suited to acute, high-arousal distress, where regulating the body first matters more than reasoning.
Hofmann et al. (2012), a review of meta-analyses spanning many disorders, supports CBT broadly. Well-supported. Cuijpers and colleagues' meta-analytic work adds an honest caveat worth taking seriously: effect sizes shrink substantially once analyses adjust for publication bias and trial quality — CBT still works, just not always to the degree older headline numbers suggested.
Behavioural activation is arguably the most useful fact here. Ekers et al.'s (2014) meta-analysis and the COBRA trial (Richards et al., Lancet, 2016) found BA delivered by non-specialist workers was non-inferior to full CBT, and cheaper. Well-supported. The least clever component of the package works about as well as all of it.
The honest complication: Jacobson et al.'s (1996) component analysis found the behavioural elements may carry much of CBT's effect, with cognitive restructuring — the Thought Record's central technique — adding less than commonly assumed. CBT holds a central place in NICE and comparable clinical guidance regardless.
The engine of CBT. Walk a hot thought through the steps and watch how much it loosens once it meets the evidence.
A thought record tests a belief on paper. A behavioural experiment tests it in the world — and testing it in the world generally moves belief further than arguing with it does.
The structure is simple: predict what will happen, and how strongly you believe it (0–10) → do the thing → record what actually happened → re-rate your belief now that you have real data.
You don't need a separate tool for this — the Thought Record above already has every field. Use 2 · Automatic thought for your prediction ("If I speak up in the meeting, everyone will think I'm wrong") and the belief slider under it for your starting confidence. Go do the thing. Come back and use 5 · Evidence FOR/AGAINST to record what actually happened, and 7 · Re-rate to see how far the real-world result moved you — often further, in one try, than a week of arguing with the thought on paper.
The ten most common thinking traps. Tap any one for a plain example — naming the trap is half of loosening it.
No time for a full record? Run the thought past four quick questions.
Familiar thoughts feel true. That's not the same thing.
Mind-reading and fortune-telling feel like knowing. They're guesses.
A thought can be true and still be worth setting down if it only drags you.
When you're low, motivation follows action, not the other way round. Plan a small activity, guess how it'll feel, do it, then check — low mood almost always under-predicts.
These are self-help skills, not therapy, and not a diagnosis. If low mood, hopelessness, or anxious thoughts are heavy or persistent — or you're thinking of harming yourself — please reach a doctor, therapist, or a crisis line. Working with a real CBT therapist is more powerful than any worksheet.
This is a self-help version of a therapy designed to be delivered by a clinician — a supplement to care, not a replacement. If distress is heavy or persistent, see Support & crisis lines or When not to use this.
Thought records are hard to use in acute distress. When arousal is high, regulate first — try TIPP or a breathing tool — and reason once things have settled.
When the thought is accurate. If a situation is genuinely bad, "testing it against the facts" can slide into invalidation. Not every hard thought is a distortion; some are just hard truths, and those call for support, not correction.
ACT skills changes your relationship to a thought instead of testing it. DBT skills is built for crisis-level emotional intensity.