Most courses of CBT run 8 to 16 weekly sessions; NICE-recommended protocols range from 6 sessions for some problems to 20+ for longstanding patterns. The honest answer for you specifically depends on the problem, how long it has been running, and what you do between sessions — here is how to estimate it.
This is one of the first questions almost everyone asks before starting therapy, and it deserves a real answer rather than “we'll see how it goes.” Therapy is a financial commitment as well as an emotional one, and CBT is unusual among therapies in that the typical duration is actually known: it is studied, published, and built into the treatment protocols themselves. So here are the typical numbers, what moves yours up or down, how progress gets reviewed along the way, and what the whole course costs.
Typical session numbers by problem
These are typical ranges, drawn from the evidence base and treatment protocols — orientation, not promises. Every person is different, and your own estimate gets much sharper after one or two sessions, once the specific pattern is mapped.
| Problem | Typical sessions | Why that range |
|---|---|---|
| Panic | 6–12 | One of the best-mapped patterns in CBT. Once the cycle of body sensation, catastrophic interpretation and avoidance is worked on, frequency tends to drop quickly. |
| Generalised anxiety / chronic worry | 8–14 | Worry is more diffuse than panic and woven into more of daily life, so the work takes somewhat longer. |
| Social anxiety | 12–16 | Involves real-world practice that has to be paced carefully, and the beliefs about being judged are usually long-standing. |
| Depression / low mood | 12–20 | Usually combines behavioural work early (doing more of what matters) with cognitive work later. Mood shifts more gradually than panic does. |
| OCD & intrusive thoughts | 12–20 | Exposure and response prevention is gradual and incremental by design — rushing it does not work. |
| Health anxiety | 10–16 | The checking and reassurance-seeking cycle responds well to structured work, but it is usually woven into daily routine and takes time to unwind. |
| Longstanding perfectionism / self-esteem | 14–20+ | The beliefs underneath have usually been operating — and even been useful — for years. Loosening them properly, rather than briefly, takes longer. |
What actually determines your number
1. How severe it is, and how long it has been running
A pattern that started six months ago has fewer layers than one that has been operating for fifteen years. Neither is harder to work with in principle, but the older, more entrenched version usually needs more sessions to genuinely shift. Severity works the same way: a problem that touches one corner of your life resolves faster than one shaping your work, relationships and sleep all at once.
2. What you do between sessions — the biggest lever
This is the single largest factor, and the one most in your control. CBT works because the work continues between sessions — the experiments, the practice, the small deliberate changes. Someone doing that consistently often finishes near the bottom of their range; someone whose weeks fill up and whose practice quietly stops tends to drift towards the top of it, or past it. When between-session work is not happening, the useful response is not guilt but a conversation: sometimes the task needs adjusting, sometimes it was badly chosen and needs replacing.
3. How clear the goal is
“I want to stop lying awake rehearsing conversations” is a faster piece of work than “I want to feel better,” because you both know what finished looks like. Part of the first session or two is turning the second kind of goal into the first kind — and that clarity alone often shaves sessions off the total.
4. One problem or several
Anxiety and low mood often arrive together; health anxiety and checking habits overlap; perfectionism is rarely the only thing going on. We work on the most pressing piece first and the others often ease alongside it, but a picture with several strands genuinely takes more sessions than a single well-defined problem — and it is more honest to say that upfront than to discover it at session ten.
How progress is reviewed — and why the aim is to finish
The number is not something I decide privately and reveal at the end. By the end of the first or second session, once the specific pattern is mapped, I can give you a realistic estimate — and then we review it openly, together, against goals that are measurable enough to argue with: sleeping through most nights, making the phone call without a script, the checking down from an hour to minutes.
I want to be direct about the stance behind this, because it is the thing I would want to know if I were choosing a therapist: the aim of CBT is to finish. Good CBT ends with you knowing your own pattern well enough to manage it yourself, not with a standing weekly appointment stretching into a second year. Therapy that quietly runs past its estimate without anyone naming it is bad practice; therapy that revisits the plan honestly when the picture changes is good practice. My incentive should be your independence, and building towards an ending — including spacing out the final sessions as you take over the work — is part of the method, not an afterthought.
What if it's not working?
Sometimes, honestly, it isn't — and the worst response is to keep going in silence and hope. If we are several sessions in and the measures have not moved, that gets said out loud and we look at why. Usually one of three things is true: the plan needs adjusting (wrong target, wrong pace, between-session work not landing); the formulation was incomplete and something bigger is underneath; or CBT is not the right approach for this problem right now, in which case the honest move is to say so and help you find something that fits better — a different kind of therapy, your GP, or NHS services. Referring on is not a failure. Quietly collecting fees for therapy that is not helping would be.
Can you do CBT faster?
Somewhat — but be suspicious of anyone promising speed. Intensive formats do exist: some clinics and research programmes deliver CBT in longer sessions over a compressed period, and for certain problems the evidence for that is reasonable. The more practical accelerator for most people is unglamorous: consistent between-session practice. Two sessions' worth of change can happen in one week when the work between them actually happens — that is what compresses timelines, not any special technique. What does not exist is a shortcut that skips the practice. If your main constraint is money rather than time, fewer sessions spaced further apart with heavier between-session work is sometimes a workable compromise — worth discussing rather than assuming.
The maths: what a whole course costs
The session number matters mostly because it turns a per-session price into a real total. At my rate of £60 a session, a typical course of 8 to 16 sessions comes to £480–£960 — a knowable, bounded figure rather than an open-ended commitment. The full breakdown of UK CBT prices, free NHS routes and what drives the differences is in how much does CBT cost in the UK? — but the short version is that a defined course with an agreed endpoint is a very different financial decision from indefinite weekly therapy, and CBT is designed to be the former.
When to seek help now rather than estimate sessions
If the problem is escalating quickly, affecting your safety, or making it hard to function day to day, the right move is support now — not comparison-shopping session counts. In England you can self-refer to NHS Talking Therapies online without going through your GP, and it is free. Elsewhere in the UK and Ireland, your GP is the fastest route into NHS or HSE services.
If you are in crisis or immediate danger, contact your GP urgently, call NHS 111, or go to A&E.
Start with a free 15-minute call
If you want a number for your situation rather than a table of ranges, that is exactly what the free call is for. I'll give you an honest range for your situation on the call — it's a fair question to ask any therapist, and one any good therapist should be willing to answer before you commit to anything. No pressure either way, and no obligation to book afterwards.
Common questions
Can CBT work in 6 sessions?
Sometimes, yes. When the problem is narrow and specific — a discrete phobia, panic that started recently, one well-defined pattern — six sessions of focused work with consistent practice can be enough, and some NICE-recommended protocols are built around roughly that number. It is the exception rather than the rule, and promising a fast finish before understanding the problem would be dishonest. What six sessions can almost always do is map the problem properly and give you a realistic estimate of what remains.
Why do some people need 20+ sessions?
Usually because the pattern is longstanding, woven into several areas of life, or more than one problem is running at once. A perfectionist pattern operating since school has more layers than panic that started three months ago — the older pattern takes more sessions to genuinely shift rather than briefly suppress. Needing 20+ is not therapy failing; it is a reason the estimate should be honest at the start and reviewed openly as you go.
Weekly or fortnightly?
Weekly, especially early on — CBT builds momentum, and fortnightly gaps at the start slow the loop of practice and review, which usually stretches the total. Fortnightly earns its place towards the end of a course, as a deliberate step-down while you take over more of the work yourself. A good way to finish; a slow way to start.
What if I stop early?
Nothing bad — therapy is voluntary and you can stop whenever you choose. Stopping because the problem has genuinely eased is a good outcome. Stopping mid-course, before new patterns bed in, means results are less likely to hold. If you are thinking of stopping, say so in a session rather than by cancelling — sometimes the plan needs adjusting, sometimes a planned earlier ending is exactly right, and both are better decided openly.
Do online courses count?
They help, but they are not interchangeable with sessions. A good self-help course or NHS digital programme teaches the same framework, and arriving already fluent in the language of CBT is a real head start. What a course cannot do is tailor the work to your specific pattern or catch where the standard advice does not fit you. If self-guided CBT resolved the problem, you do not need me. If it helped but the problem is still there, the work usually needs personalising — it does not mean CBT is not for you.
MSc Psychological Therapies (University of Edinburgh). I practise evidence-based CBT online across the UK & Ireland, and research what makes therapy work. More about me · Book a free call