How Care Works

What does CBT actually look like between sessions?

How cognitive behavioral therapy turns insight into practice, what useful between-session work can include, and what to do when therapy homework feels like too much.

Dr. Ramy ElsawahPsychiatrist & FounderUpdated September 20267 min read
An adult writes a small practice plan in a notebook at a kitchen table
Key points
  • CBT often uses small between-session experiments so skills meet real life instead of staying in the therapy room.
  • Good practice is collaborative, specific, tied to your goals, and reviewed without shame.
  • A thought record is one option, not the universal uniform of CBT.
  • If an assignment feels unsafe, confusing, culturally off, or impossible, that reaction belongs in the next session.

Your therapy session ends with a useful realization and a small plan. By Tuesday, the worksheet is under a grocery receipt, the week has become feral, and the phrase “therapy homework” is producing the exact avoidance response it was meant to study.

Cognitive behavioral therapy, or CBT, is a family of structured therapies that looks at connections among thoughts, emotions, body sensations, and actions. It’s often active. That can include practicing a skill between sessions, but useful practice isn’t punishment, busywork, or a gold-star system for being psychologically organized. You’re not being graded.

The week is where the treatment gets tested

A therapy hour can help you notice a pattern. Real life shows whether a new response is usable when your boss sends the email, your heart races in the grocery line, or the dishes become a referendum on your relationship. Between-session practice carries one idea out of the office and into the situation that keeps teaching the old lesson. It won’t be tidy, and that’s useful.

Research on CBT homework generally finds that greater engagement is associated with better outcomes. That doesn’t prove that completing more boxes automatically causes recovery. People who are improving may find practice easier, studies measure completion differently, and the quality of the work matters. The sensible conclusion is that practice can help when it fits the treatment and the person.

I don’t think a good assignment should arrive like a surprise tax form. You and the therapist should understand why you’re doing it, what it’s meant to test, and how small it can be while still teaching something. If neither of you can explain the point, the worksheet may be the confused party.

Practice can be behavioral, cognitive, or both

Sometimes you track a situation, the thought that appeared, the emotion that followed, and what you did next. The goal isn’t to replace every difficult thought with a cheerful slogan. It’s to slow the sequence enough to examine whether the thought is accurate, useful, incomplete, or being treated as a fact. You don’t have to admire the thought before questioning it.

Sometimes the work is behavioral. You might schedule one activity when depression has narrowed the week, approach a feared situation in planned steps, practice an assertive request, change a sleep habit, or test a prediction. If your mind says, “If I ask one question, everyone will think I’m incompetent,” an experiment collects better evidence than an hour of internal litigation.

Other assignments build awareness: noting panic cues, catching reassurance seeking, recording sleep timing, or noticing when self-criticism changes behavior. Tracking should answer a question. If it turns your day into a surveillance project, bring that back to therapy.

A useful experiment has a job description

Before leaving the session, you should know the task, the situation, the frequency, and the reason. “Be less avoidant” isn’t an assignment. “On Wednesday, open the bill for two minutes and write down the amount before deciding the next step” is specific enough to attempt. If you can’t picture when it happens, it probably isn’t ready.

Predict what you think will happen. Rate how strongly you believe it if that’s useful. Afterward, record what actually happened and what you learned. The experiment can still be valuable if you felt anxious, the conversation was awkward, or you stopped early. CBT isn’t auditioning you for the role of Person Who Never Struggles.

Keep the dose realistic. A five-minute practice completed in the right situation may teach more than a beautiful 45-minute worksheet finished at midnight out of guilt. A 2023 systematic review found that collaborative planning, a clear rationale, written summaries, flexibility, and reviewing the work can support engagement. The relationship around the assignment matters.

Not doing it is information, not a confession

If you didn’t practice, start with what got in the way. Did you forget? Was the task too vague, too large, embarrassing, physically inaccessible, or emotionally overwhelming? Did it clash with work, caregiving, culture, safety, or what you actually want from therapy? Those answers improve the plan. They don’t cancel it.

A therapist shouldn’t use homework to shame you, prove resistance, or avoid understanding your circumstances. You shouldn’t have to pretend completion to protect the therapist’s feelings. A missed assignment can reveal that the cue was weak, the step was too steep, or the rationale never landed.

Try a repair question: “What is the smallest version that would still teach us something?” Put the practice beside an existing cue, such as after brushing your teeth or before opening email. Ask for a written note if you’ll forget. If forms make you freeze, use a voice memo or a single sentence. The method serves the learning, not the other way around.

Exposure is planned practice, not surprise flooding

For anxiety and obsessive-compulsive problems, CBT may include exposure: approaching feared situations, sensations, memories, or uncertainty in a deliberate way while reducing avoidance or rituals. It should have a rationale, a plan, and consent. It isn’t a therapist tricking you into the hardest possible task, and it shouldn’t be a surprise.

The step should be challenging enough to teach, not so large that the only lesson is “never agree to this again.” You may practice staying in a situation long enough to learn that anxiety changes on its own, that uncertainty can be tolerated, or that a feared outcome is less likely or more manageable than predicted. The learning goal matters more than looking calm.

If a task raises concerns about physical safety, trauma, substance withdrawal, eating-disorder risk, compulsions, or another medical issue, don’t improvise. Pause and contact the treating clinician. General worksheets can’t decide what exposure is appropriate for you.

You should be able to see the thread

At the next session, review what happened. What did you notice? What surprised you? What needs adjusting? If assignments are repeatedly given and never discussed, they can start feeling like paperwork placed into the ocean. Review connects the practice to the treatment plan, so it doesn’t disappear.

Ask how the work relates to your goals and how progress will be noticed. CBT can be adapted, and it isn’t the only effective therapy. If the approach feels rigid, invalidating, or mismatched after honest discussion, that may be a fit problem rather than a personal failure.

Try one small experiment today: write down a recurring prediction, then name one safe action that could gather evidence. You don’t have to perform the action yet. Making the question testable is already different from letting the prediction run the meeting.

If therapy practice brings a sharp increase in suicidal thinking, an inability to stay safe, or another crisis, contact your clinician or crisis support rather than waiting for the next appointment. In the United States, call or text 988. Use 911 for immediate life-threatening danger.

The bottom line: CBT between sessions should be a small, understandable experiment tied to your goals. It can involve noticing patterns, testing predictions, changing behavior, or practicing exposure. When the task doesn’t fit, say so. Good CBT uses the obstacle as information and redesigns the practice with you.

Sources: National Institute of Mental Health, Psychotherapies; Kazantzis and colleagues, CBT homework quality and outcome meta-analysis, Behavior Therapy (2016); Mausbach and colleagues, homework compliance and therapy outcomes meta-analysis, Cognitive Therapy and Research (2010); Bryant and colleagues, integrating between-session homework systematic review, Psychotherapy (2023).

This is general education, not medical advice. It cannot prescribe an exposure or determine whether a CBT assignment is safe or appropriate for you. Discuss concerns, worsening symptoms, and treatment fit with your clinician. For a mental health crisis in the United States, call or text 988; call 911 for a life-threatening emergency.
A thoughtful next step

Wondering what kind of treatment would fit?

A free 15-minute intro call can help you decide whether psychiatric care fits the next step.

Book a free 15-minute intro call