How to Tell If CBT for Depression Is Working
Last reviewed: 08/30/2026
Reviewed by: Dr. Kiesa Kelly
A few weeks into therapy, most people hit the same question: is this actually helping, or am I just talking? It is an uncomfortable question to sit with, and it is easy to answer badly in either direction — quitting something that was starting to work, or staying for months in something that was not.
Depression itself makes this harder. It flattens your sense of change, pulls memory toward the worst recent day, and makes "a little better" feel indistinguishable from "no different." So the honest answer is that you should not have to judge this from feel alone. Cognitive behavioral therapy for depression is designed to be measured, and a good course of it gives you something more reliable to look at than your mood on the morning you happen to ask.
In this article, you'll learn:
What actually gets tracked in CBT for depression, beyond "how was your week"
How to read a change in your scores without over-reading a single bad week
When it is fair to expect movement, and when it is genuinely too early to tell
What a plateau usually means, and what typically happens next
Concrete questions to bring to your clinician about your own results
The short answer: how progress is tracked
Progress in CBT for depression is tracked two ways at once — a repeated symptom measure, and a running picture of how your life is actually going. Neither one is sufficient alone.
The symptom side is usually a short questionnaire you complete before sessions. The most common one is the PHQ-9, a nine-item self-report scored 0 to 27, which was validated as a measure of depression severity rather than only as a yes-or-no screen [1]. That distinction is the whole reason it works for tracking: it was built to move as you move.
Using those repeated scores deliberately — reviewing them with the client, in session, and adjusting the plan when they stall — is called measurement-based care. It is not just record-keeping. A systematic review of randomized trials found that measurement-based care improved symptom outcomes and remission rates for depressive disorders compared with usual care [2], and it is recommended as a principle of depression management in current clinical guidelines [3].
📈 Key takeaway: The score is a tool for the conversation, not a grade. Its value comes from being looked at with your clinician, repeatedly, and acted on.
Three beliefs get in the way here, and they are worth naming before we go further.
"If I still feel bad, nothing is working." Recovery from depression is usually not a smooth downhill line. Sleep and appetite often shift before mood does, and energy often returns before interest does. It is entirely normal to be measurably better and still feel bad on the day you are asked.
"The questionnaire is just paperwork." It is the opposite. Filling it out casually, or filling it out to look better than you feel, removes the one instrument that can see change you cannot. Under-reporting to avoid disappointing your therapist is common and it costs you accuracy.
"A good therapist should just be able to tell." No clinician can reliably detect a modest change in someone else's internal state week to week — which is exactly why the field moved toward structured measurement in the first place. A therapist who tracks is not being impersonal; they are refusing to guess.
What gets measured
Symptom-level change
The PHQ-9 asks about the nine symptom areas used to define depression, each rated from "not at all" to "nearly every day" over the past two weeks. Scores of roughly 5, 10, 15, and 20 mark the boundaries between mild, moderate, moderately severe, and severe [1]. Large pooled analyses of individual patient data have since confirmed that a cutoff around 10 performs best when the questionnaire is used for detection [4], a finding that held in an updated analysis drawing on a larger pooled sample [10].
For tracking, two benchmarks come up most often. A change of about 5 points is generally treated as a meaningful shift for an individual, and a 50% reduction from your starting score is the conventional definition of treatment response [5]. Both are useful anchors. Neither is a pass mark, and more recent work suggests the size of a genuinely meaningful change depends partly on how severe things were at the start.
Here is what that looks like in practice. Suppose you started at 19 — moderately severe. Four weeks in you are at 16, and you feel like nothing has changed. On paper that is a 3-point drop, short of the usual benchmark, and easy to read as failure. But you have started sleeping through more nights, you answered two work emails you had been avoiding for a month, and you noticed yourself irritated rather than numb. That is early movement in the symptoms that typically shift first, and it is a reasonable basis for staying the course another few weeks rather than concluding the approach is wrong.
Or: you started at 14 and you are now at 8. The number says you have responded well. But you are still not returning calls from friends, you are still eating lunch alone in your car to avoid the break room, and the drop came almost entirely from the sleep and appetite items. That is real improvement and an incomplete one, and it is worth saying out loud rather than letting the score close the topic.
Many people are also carrying anxiety alongside depression, and the two do not always move together. When that is part of the picture, clinicians often track a second brief measure such as the GAD-7 so a stall in one is not hidden by progress in the other. You can look at the full set of screeners we use if you want to see what these instruments actually ask.
🧭 Key takeaway: Ask which items moved, not just whether the total moved. Two people with the same 6-point drop can be in very different places.
Day-to-day functioning
The second half is the part a questionnaire cannot capture. CBT for depression is built around changing what you do — behavioral activation, scheduling activity, testing the predictions your low mood makes about how things will go. So the functional questions are direct: are you getting out of bed closer to when you intend to? Are you doing any of the things that used to matter to you? Has anyone close to you noticed a difference?
This matters because functioning and symptom scores can move out of step in both directions. Someone can post a much lower score while their life has not changed shape at all, which usually means the gains are fragile. And someone can show only a modest score change while going back to work, resuming a routine, and re-entering their own life — which is often the more durable result.
🔎 Key takeaway: If your score improved but your week looks identical, that is worth raising. It usually means something is being measured that has not yet been lived.
How often it is reviewed
In most structured CBT, the measure is completed before each session and reviewed in the session — that in-session review is what separates measurement-based care from filing paperwork [2]. Weekly data points give you a trend line rather than two isolated snapshots, and a trend is the only thing that can distinguish a bad week from a stalled course.
The timeline is shorter than many people assume. NICE guidance describes individual CBT for depression as usually running about eight regular sessions, with additional sessions where there are co-occurring conditions, complex circumstances, or leftover symptoms to address [6]. Courses in US practice often run longer, but the underlying point holds: this is a treatment with a defined arc, not an open-ended arrangement, and there should be a point where someone says out loud whether it is working. If you want a fuller picture of typical timelines, we cover that in how long depression treatment takes.
A fair trial usually means several weeks of consistent attendance and consistent between-session work. CBT for depression asks you to do things between appointments, and a course where the homework has not happened has not really been tested yet. That is not a criticism — difficulty initiating tasks is a symptom, not a character flaw — but it does change how you interpret a flat line.
⏳ Key takeaway: One flat check-in means very little. Two or three in a row is a signal, and it is the signal worth acting on.
What a plateau means and what usually follows
Plateaus are common and they are not the same thing as failure. Scores frequently drop quickly at first and then flatten well short of where you want to be. What matters is what happens next, because a plateau is information about the plan rather than a verdict on you.
Usually one of a few things is going on. The work may have drifted away from what is actually maintaining the depression — the sessions are useful conversations, but the behavioral piece has quietly fallen away. Something untreated may be holding the floor up: unaddressed anxiety, chronic insomnia, unprocessed trauma, or a physical health issue with overlapping symptoms. Life circumstances may be actively producing the depression, in which case symptom-focused work alone will keep hitting a ceiling. Or the intensity may simply be wrong, and the honest answer is more support rather than a different flavor of it.
Here is the decision heuristic worth carrying out of this article:
If your scores are moving but slowly, and your functioning is moving too — continue and reassess in three or four weeks. This is what ordinary recovery looks like.
If your scores are flat but your functioning is genuinely improving — continue, and say so; the measure may be missing what is changing.
If both are flat across two or three consecutive reviews — raise it directly. This is the point to change something, not to try harder at the same thing.
If things are getting worse, or you are having thoughts of harming yourself — say that now, at the next contact, not at the next scheduled review.
What "changing something" looks like is usually specific rather than dramatic: sharpening the focus of the work, adding behavioral activation if the sessions have gone mostly cognitive, treating a co-occurring problem that is holding things down, adding a medication consultation, or moving to a different modality. CBT itself rests on a large evidence base here — a meta-analysis pooling 409 trials and more than 52,000 patients found it effective for depression across different patient groups and settings [8] — and it is one of several approaches that professional practice guidelines endorse for this condition [9]. Switching between them is a normal clinical move rather than a restart, and we walk through the main options in choosing between CBT, ACT, and DBT for depression.
Format can matter here too. Some people plateau in telehealth and move again once sessions are in a room, and for others the reverse is true. If you are weighing that, our page on CBT for depression in Nashville lays out how the in-person option works alongside telehealth across Tennessee.
🔁 Key takeaway: A plateau is a prompt to change the plan. Staying in something unchanged for months because it once helped is the most common way a good treatment goes stale.
Talking with your clinician about the results
The single most useful thing you can do is ask to see the trend. Not this week's number — the line. Most clinicians who use these measures can show you the sequence, and looking at it together tends to resolve the "am I imagining this" question quickly in one direction or the other.
Questions worth asking, close to verbatim:
"Can we look at my scores from the start until now, together?" You are asking for the trend, not the snapshot.
"Which specific items have moved, and which haven't?" This is where a flat-looking total often turns out to be hiding real movement, or vice versa.
"What would you expect to see change next, and by when?" A clear answer gives you a checkpoint. A vague one is itself worth noticing.
"If this is still flat in three weeks, what would you change?" Asking before the plateau makes the conversation easier than asking during it.
"Is there something else going on that we haven't addressed?" Co-occurring anxiety, sleep problems, or trauma frequently set the ceiling on depression treatment.
If raising this feels like criticizing your therapist, it is worth knowing that from the clinician's side it is close to the opposite. A client who says "I don't think this is moving" is giving you the most useful information available, and it is a normal part of collaborative care rather than a complaint. Our clinicians work this way across our specialized therapy services, and the review conversation is built into the work rather than bolted on at the end.
One thing that is not up for scheduling: if you are having thoughts of hurting yourself, that belongs in the conversation immediately, not at the next review. Item 9 of the PHQ-9 asks about exactly this, and a nonzero answer there is always worth naming out loud. If you are in crisis or thinking about suicide, you can call or text 988 in the United States to reach the Suicide & Crisis Lifeline, any time [7].
🤝 Key takeaway: Saying "I don't think this is working" is participating in your treatment, not quitting it.
Next step: getting support
You do not have to work out from feel alone whether therapy is helping. Between a repeated measure, an honest look at how your weeks are actually going, and a clinician willing to say what they expect to change and by when, the question becomes answerable — usually well before you have spent months wondering.
If you are already in treatment, the next step is small: ask to see your trend, and ask what happens if it stays flat. If you are not yet in treatment, the useful thing to look for is not a promise about outcomes but a clear account of how progress will be tracked and reviewed.
Feeling weighed down lately?
Depression is treatable, and the right support makes a difference — a clinician can help you understand what's going on and what would help you feel like yourself again.
Frequently Asked Questions
How much does a PHQ-9 score need to drop to count as real progress?
Two benchmarks are commonly used. A drop of about 5 points is generally treated as a meaningful change for one person, and a 50% reduction from where you started is the usual definition of treatment response. Both are rules of thumb rather than thresholds you pass or fail. A smaller drop paired with real change in sleep, work, or relationships still counts, and your clinician reads the number alongside how your weeks are actually going.
What type of therapist is best for depression?
Look for training in an approach with evidence behind it for depression — cognitive behavioral therapy, behavioral activation, and interpersonal therapy all have strong research support. The specific license matters less than whether the clinician works from a structured method and tracks progress with you over time. Ask directly how they will know whether treatment is working; a clinician with a clear answer to that question is usually a good sign.
How to know if therapy is not working?
The clearest signal is no measurable movement after a fair trial — several weeks of consistent attendance and between-session work, with both your scores and your daily functioning staying flat. One flat check-in is not enough to judge, because normal weeks vary. If two or three reviews in a row show nothing shifting, that is the point to raise it with your clinician rather than quietly waiting it out.
How to treat depression and anxiety without medication?
Psychotherapy on its own is an evidence-based option for many people, and major clinical guidelines list it as a first-line choice. CBT, behavioral activation, and exposure-based work for anxiety all have research support without medication. More severe depression often responds better to therapy and medication together, so it is worth treating that as a decision to make with a prescriber and a therapist rather than one to rule out in advance.
About the Author
Dr. Kiesa Kelly is a licensed clinical psychologist and the founder of ScienceWorks Behavioral Healthcare, with more than 20 years of experience in psychological assessment and evidence-based treatment. Her background includes clinical training at the University of Chicago, Vanderbilt University, and the University of Wisconsin, along with NIH-funded research training.
Her work centers on structured, measurable approaches to mood and anxiety conditions — the kind of treatment where progress is tracked openly with the client rather than inferred. She reviews every clinical article published on this site for accuracy before it goes live.
References
1. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: validity of a brief depression severity measure. Journal of General Internal Medicine. 2001;16(9):606–613. https://pmc.ncbi.nlm.nih.gov/articles/PMC1495268/
2. Guo T, Xiang Y-T, Xiao L, et al. The efficacy of measurement-based care for depressive disorders: systematic review and meta-analysis of randomized controlled trials. Journal of Clinical Psychiatry. 2021. https://pubmed.ncbi.nlm.nih.gov/34587377/
3. Lam RW, Kennedy SH, Adams C, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) 2023 update on clinical guidelines for management of major depressive disorder in adults. Canadian Journal of Psychiatry. 2024. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11351064/
4. Levis B, Benedetti A, Thombs BD. Accuracy of Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ. 2019;365:l1476. https://pubmed.ncbi.nlm.nih.gov/30967483/
5. Kroenke K, Spitzer RL, Williams JBW, Löwe B. Monitoring depression treatment outcomes with the Patient Health Questionnaire-9. Medical Care. 2004;42(12):1194–1201. https://pubmed.ncbi.nlm.nih.gov/15550799/
6. National Institute for Health and Care Excellence. Depression in adults: treatment and management. NICE guideline NG222. 2022. https://www.nice.org.uk/guidance/ng222
7. 988 Suicide & Crisis Lifeline. https://988lifeline.org/
8. Cuijpers P, Miguel C, Harrer M, et al. Cognitive behavior therapy vs. control conditions, other psychotherapies, pharmacotherapies and combined treatment for depression: a comprehensive meta-analysis including 409 trials with 52,702 patients. World Psychiatry. 2023;22(1):105–115. https://doi.org/10.1002/wps.21069
9. American Psychological Association. Clinical practice guideline for the treatment of depression across three age cohorts. 2019. https://www.apa.org/depression-guideline
10. Negeri ZF, Levis B, Sun Y, et al. Accuracy of the Patient Health Questionnaire-9 for screening to detect major depression: updated systematic review and individual participant data meta-analysis. BMJ. 2021;375:n2183. https://pubmed.ncbi.nlm.nih.gov/34610915/
Disclaimer
This article is for informational purposes only and is not a substitute for individualized clinical assessment, diagnosis, or treatment. Screening questionnaires such as the PHQ-9 and GAD-7 are tools for tracking and discussion, not diagnostic tests, and a score cannot confirm or rule out a condition on its own. If you are experiencing thoughts of harming yourself, contact the 988 Suicide & Crisis Lifeline or your local emergency services.