The short answer: Give therapy a fair trial, then review it out loud instead of quietly quitting. Pooled data on session counts suggest about half of people show measurable improvement by 8 sessions. If nothing has moved after 4 to 6 weeks, guidelines tell clinicians to recheck the diagnosis, look for other conditions, and change the plan. Switching therapist, switching therapy type, adding medication, or asking for a specialist referral are all standard next steps.
How long should therapy take before it helps?
The best known estimate comes from a 1986 analysis in American Psychologist by Howard and colleagues. They pooled 15 samples covering 2,431 outpatients in weekly individual psychotherapy. About 50% were measurably improved by 8 sessions, about 75% by 26 sessions, and about 85% by the end of a year.
Two details matter more than the headline. First, the curve is front loaded. Most of the change happens early, and each extra session adds less than the one before it. Second, depression responded at the lowest doses in that analysis. For depression and anxiety, half of patients improved in roughly 8 to 13 sessions.
Guidelines use a shorter window before a formal review. NICE guideline NG222 says that if depression has not responded at all after 4 to 6 weeks of psychological therapy, the clinician should sit down and work out why.
These are averages across large groups, not a schedule for one person. They are useful for one purpose: telling the difference between "too early to judge" and "this has stalled."
Is your progress actually being measured?
Without a number, "therapy is not helping" is a feeling that you and your therapist can honestly disagree about. Measurement-based care fixes that. The 2022 VA/DoD guideline for major depressive disorder describes three steps: collect a validated measure, share the result with the patient, and act on it. Its Recommendation 2 suggests using a quantitative severity measure at the start and at regular intervals after that.
The evidence is real but uneven. A 2015 randomized trial in the American Journal of Psychiatry assigned 120 outpatients with moderate to severe depression to measurement-based care or standard care for 24 weeks. Response was 86.9% against 62.7%, and remission was 73.8% against 28.8%. That trial tested medication management, not talk therapy.
For psychological treatment the effect is smaller. A 2021 multilevel meta-analysis of 58 studies and 21,699 patients found progress feedback produced a small gain in symptom reduction (d = 0.15) and a modest drop in dropout. Small is not nothing when the alternative is guessing.
A practical ask: complete a short scale such as the PHQ-9 at intervals, and look at the trend together.
How much does the fit with your therapist matter?
A great deal. The therapeutic alliance means three things: agreement on goals, agreement on the tasks of therapy, and the emotional bond. A 2018 meta-analysis of 295 studies and more than 30,000 patients found an alliance to outcome correlation of r = .278. The link held across treatment approaches, countries, and who did the rating.
A poor fit often shows up as a rupture: a disagreement about goals, a lack of collaboration, or a strain in the bond. Ruptures are common and are not automatically a reason to leave. A companion meta-analysis of 11 studies and 1,314 patients found that repairing a rupture was moderately associated with better outcomes (r = .29).
That finding has a blunt practical meaning. Saying "I do not think this is working, and here is why" is not rude. It is often the intervention.
Does the type of therapy matter?
Less than most people expect. A 2013 network meta-analysis in PLOS Medicine pooled 198 studies and 15,118 adults with depression. All seven psychotherapies beat a waitlist control with moderate to large effects. Differences between the therapies themselves were absent to small.
The VA/DoD guideline reflects that. It lists acceptance and commitment therapy, behavioral activation, cognitive behavioral therapy, interpersonal therapy, mindfulness-based cognitive therapy, problem-solving therapy, and short-term psychodynamic therapy as options, not ranked against each other.
So swapping CBT for something else is not a guaranteed fix. It can still help when the current focus does not match the problem. A person stuck in withdrawal and inactivity may need a behavioral focus. Someone whose depression tracks a grief or a role change may do better with an interpersonal focus. Our guide to depression treatment options compared walks through what each approach targets.
Could something else be driving the depression?
Therapy can stall because the target is wrong. NICE recommendation 1.9.2 says that when depression has not responded, clinicians should review the diagnosis and consider alternative or comorbid conditions. The VA/DoD guideline says much the same: reassess for possible diagnostic error, co-occurring conditions, and treatment adherence.
Common things that get missed:
- Bipolar disorder. A 2011 primary care screening study in the British Journal of Psychiatry estimated that between 3.3% and 21.6% of patients treated for unipolar depression may have an undiagnosed bipolar disorder. The range is wide because the definition used changes the answer. If you have had periods of unusually elevated mood, reduced need for sleep, or fast thinking, say so. Bipolar.md covers how that assessment works.
- Trauma, ADHD, and substance use. Guidelines treat these as co-occurring conditions to screen for, because untreated they can hold depression in place.
- Medical causes. The VA/DoD guideline lists ruling out depression secondary to other causes, including hypothyroidism and vitamin B-12 deficiency. See our review of vitamin D and B12 deficiency and depression.
- Sleep apnea. In a CDC analysis of 9,714 US adults, diagnosed sleep apnea was associated with probable major depression (odds ratio 2.4 in men and 5.2 in women).
What are the next steps when therapy is not working?
NICE sets out an order. First, discuss whether personal, social, or environmental factors explain the lack of progress, and whether there have been problems sticking to the plan, such as missed sessions. Then review the diagnosis. Then change the treatment. For a limited response to psychological therapy alone, the listed options are switching to a different psychological treatment, adding an SSRI, or switching to an SSRI alone.
The VA/DoD guideline offers a similar menu: add psychotherapy or an antidepressant, switch between the two, switch to a different focus of psychotherapy, or augment with a different class of medication.
Persistence is worth something. In the STAR*D trial, 36.8% of patients reached remission on the first treatment step, and the cumulative remission rate after four steps was 67%. Later steps had lower remission rates and higher relapse rates, so more steps is not a free ride.
Before you quit, consider raising these with your therapist:
- Ask for one written goal and one measure, reviewed on a set date.
- Say plainly what is not working, including about the relationship itself.
- Ask what the plan is if the score has not moved by that date.
- Ask for a diagnostic review, including screening for bipolar symptoms, trauma, ADHD, and substance use.
- Ask your primary care clinician about thyroid, B12, and sleep testing.
- Ask whether adding medication, or changing therapy focus, is the better next move.
- Ask for a referral if the answer needs a specialist.
If two or more adequate treatments have failed, that pattern has a name and its own pathway. Read more on treatment-resistant depression.
If you are thinking about harming yourself, do not wait for the next appointment. In the US you can call or text 988, or chat at 988lifeline.org. The National Institute of Mental Health lists it as the national crisis line. It is free and available 24 hours a day.
The bottom line
Therapy that is not helping is a signal to review, not automatically a signal to stop. Check the dose against the evidence: roughly half of people improve by 8 sessions, and guidelines review a total non-response at 4 to 6 weeks. Then check the three things most likely to be wrong: whether progress is being measured, whether the working relationship is sound, and whether the diagnosis is complete. Changing therapist, changing approach, adding medication, and asking for a specialist referral are all ordinary parts of depression care, not failures.
Last updated: September 2026. This article is for informational purposes only and does not constitute medical advice. Decisions about starting, changing, or stopping therapy or medication for depression should be made with a qualified clinician who knows your history.