CBT session notes have more moving parts than general therapy notes — automatic thoughts to capture verbatim, homework to track, scale scores to record, and a session structure that repeats predictably enough that deviations are clinically meaningful.

The five examples below cover the most common presentations in CBT practice: depression with behavioral activation, generalized anxiety, OCD with ERP, social anxiety, and a session where progress has stalled. Each includes clinical commentary on what the note is doing and why.

What every CBT session note needs

Before the examples: here’s the core structure that works across all presentations.

Date:                    Session #:
Client (initials):       Duration:

AGENDA
What was planned:

HOMEWORK REVIEW
Assigned:
Completed: yes / partially / no
Clinical note on completion:

MAIN WORK
Automatic thought (verbatim):
Situation:
Emotion + intensity (0–10):
Distortion:
Alternative thought:
Believability of alternative (0–100%):

SCALES (if today)
PHQ-9:        GAD-7:        Other:

RISK ASSESSMENT
Suicidal ideation:

NEW HOMEWORK
Task:
Success criterion:

NEXT SESSION
Focus:

Example 1: Depression with behavioral activation

Date: 05/21/2026 | Session: 8 | Client: L.H. | Duration: 50 min

Agenda: review week’s activity log; continue behavioral activation; introduce thought monitoring

Homework review: Assigned: complete activity log daily, rate mood 0–10 after each activity. Completed: 5 of 7 days. Did not complete Saturday or Sunday. Clinical note: completion on weekdays suggests structure is a maintaining factor. Weekend drop-off consistent with “nothing matters on weekends” cognition flagged in session 6 — address today.

Main work: Explored the weekend avoidance. Client identified: “I don’t deserve to enjoy anything when I haven’t been productive.”

  • Situation: Saturday afternoon, nothing scheduled
  • Automatic thought: “What’s the point of doing anything? It won’t help.”
  • Emotion: hopelessness (8/10), lethargy (9/10)
  • Distortion: all-or-nothing thinking, disqualifying the positive
  • Alternative: “Doing one thing I used to enjoy doesn’t require me to feel productive first.” Believability: 35/100 (“I understand it but it feels hollow”)
  • Reviewed activity log: mood consistently 2–3 points higher after any activity vs. before — used as behavioral evidence against “it won’t help”

Scales: PHQ-9 = 14 (was 19 at intake, session 1). Positive trajectory noted.

Risk: No SI — confirmed by direct inquiry. Protective factors stable (living with partner, engaged in treatment).

Homework: Task: one planned activity each day this weekend (client’s choice, minimum 20 minutes). Rate mood before and after. Success criterion: complete both days; bring completed ratings to next session.

Next session: Review weekend data. If mood-activity correlation holds, use as evidence to challenge “it won’t help.” Introduce thought record format.

Clinical commentary: The homework completion pattern is documented with its clinical meaning — not just “partially completed.” The automatic thought is recorded verbatim, not paraphrased. The alternative thought’s believability (35/100) is captured because it’s clinically more useful than knowing whether restructuring was “done.”

Example 2: Generalized anxiety — worry and avoidance

Date: 05/21/2026 | Session: 6 | Client: M.K. | Duration: 50 min

Agenda: review worry postponement technique; address manager avoidance; begin exposure hierarchy

Homework review: Assigned: use worry postponement (schedule 15-minute “worry time” daily); note avoidance urges. Completed: used worry postponement 4 of 7 days. Did not schedule the manager check-in (third consecutive week). Clinical note: partial success with worry postponement — good. Persistent manager avoidance now a clear treatment target; avoidance is maintaining the anxiety cycle.

Main work: Reviewed avoidance-anxiety cycle with specific reference to manager avoidance. Client recognized the pattern (“avoiding it makes me more anxious — I know this, but I still avoid”).

  • Situation: anticipating the manager check-in
  • Automatic thought: “She’s going to think I don’t know what I’m doing.”
  • Emotion: anxiety (8/10), shame (6/10)
  • Distortion: mind-reading, probability overestimation
  • Alternative: “I don’t have evidence she thinks this. She assigned me this role.” Believability: 40/100

Introduced exposure hierarchy. Client identified 5 steps for manager interaction; rated scheduling the check-in at 5/10 distress (not the meeting itself — just scheduling it). Agreed this is a workable starting point.

Scales: GAD-7 = 13 (was 17 at intake). Improving.

Risk: No SI.

Homework: Task: schedule (not conduct) the manager check-in before next session. Note the automatic thoughts when about to do it. Success criterion: calendar invite sent. Does not need to have occurred.

Next session: Review scheduling outcome. If completed: process automatic thoughts noted. If not: explore barriers, revise hierarchy.

Clinical commentary: The “third consecutive week” detail matters — it documents a pattern, not a one-time occurrence. The homework is explicitly low-barrier: scheduling, not the meeting. The success criterion is binary and verifiable.

Example 3: OCD with ERP

Date: 05/21/2026 | Session: 10 | Client: R.T. | Duration: 50 min

Agenda: review contamination ERP; address in-session reassurance-seeking; advance hierarchy

Homework review: Assigned: touch doorknob without handwashing (first contamination hierarchy step, rated 4/10); two checking exposures (consolidation). Completed: checking exposures — both done, both successful (“it was hard but I didn’t go back”). Contamination exposure — attempted once, completed, “felt disgusting for 20 minutes then it passed.” Clinical note: successful completion of first contamination step is significant. The distress duration (20 min) and habituation are important to document for client psychoeducation.

Main work: Reviewed contamination exposure in detail. Client described urge to wash as “10/10 at the start, then it just… dropped.” Used this as in-vivo evidence for ERP rationale: distress is temporary and habituates without compulsion.

In-session reassurance-seeking noted: client asked three times whether “doing it right” would accelerate improvement. Named the behavior directly: “I notice you’re asking me to reassure you — that’s a mental compulsion in the same family as the physical ones. What happens if I don’t answer?” Client able to tolerate uncertainty after brief exploration.

Y-BOCS self-report: 19 (was 28 at intake, session 1).

Risk: No SI.

Homework:

  1. Contamination exposure: advance to touching shared office surfaces (rated 6/10). No handwashing for 2 hours after.
  2. Checking consolidation: continue 2 per week.
  3. Reassurance practice: notice reassurance-seeking urges (from others) and delay response by 10 minutes. Success criterion: attempt all three; bring urge ratings.

Next session: Review contamination exposure; introduce mental compulsion ERP if today’s reassurance work is confirmed.

Clinical commentary: The reassurance-seeking is documented as a clinical event, not overlooked. The habituation data (distress 10 → habituation over 20 min) is recorded because it becomes evidence for the ERP rationale in later sessions. Y-BOCS trajectory referenced from intake.

Example 4: Social anxiety

Date: 05/21/2026 | Session: 9 | Client: A.M. | Duration: 50 min

Agenda: post-exposure debrief (last week’s presentation); update social anxiety hierarchy; introduce video feedback if ready

Homework review: Assigned: give the work presentation (planned exposure, rated 7/10 distress). Completed: yes — “I did it. I was shaking the whole time but I did it.” Clinical note: completion of highest-rated exposure to date. First successful high-stakes public speaking exposure in treatment. Key moment to consolidate.

Main work: Detailed debrief of presentation experience.

  • Pre-exposure prediction: “They’ll see I’m nervous and lose confidence in me.” Confidence in prediction: 85/100.
  • Actual outcome: received positive feedback from two colleagues. Manager nodded throughout.
  • Post-exposure prediction accuracy: “I was more nervous than I looked. Nobody mentioned it.”
  • Revised prediction confidence: 30/100.

Introduced video feedback rationale: watching oneself present as disconfirmatory evidence for “I look as nervous as I feel.” Client ambivalent but agreed to try with a low-stakes recording (talking to camera alone, not at work).

Safety behaviors identified still in use: gripping notes throughout presentation, avoiding eye contact with specific colleague. Added to hierarchy as future targets.

Scales: LSAS (Liebowitz Social Anxiety Scale) sub-administered (5 items): mild improvement.

Risk: No SI.

Homework: Record a 3-minute video of yourself presenting something (topic: your choice). Watch it once. Rate predicted vs. actual “how nervous I look” before and after. Success criterion: record and watch at least once.

Next session: Review video feedback. If successful disconfirmation: process. Introduce eye contact fading as next hierarchy step.

Clinical commentary: The pre/post prediction format is the structural core of cognitive work in social anxiety CBT. Recording it creates data that can be reviewed across sessions. Safety behaviors are noted as future hierarchy targets, not just observed.

Example 5: Progress plateau — what the note looks like

Date: 05/21/2026 | Session: 16 | Client: D.V. | Duration: 50 min

Agenda: review progress; explore plateau; adjust treatment plan

Homework review: Assigned: thought records (3 per week). Completed: 1 record completed. Client: “I keep forgetting, and honestly I’m not sure it’s helping anymore.” Clinical note: homework completion has declined over the last 4 sessions (from 7/7 to 5/7 to 3/7 to 1/7). This is a clinical pattern, not a motivational problem to be addressed with encouragement alone.

Main work: Named the pattern directly: “I’ve noticed your homework completion has been declining, and I want to understand what’s happening — not push you harder on it.”

Client expressed: “I feel like we’ve been doing the same thing for weeks and I’m stuck at the same place.”

Explored: PHQ-9 has been 10–12 for the last 5 sessions (was 19 at intake). Partial response, not full remission. Client’s subjective sense of “stuck” matches the scale trajectory.

Hypotheses explored jointly:

  1. Thought records are not targeting the maintaining mechanism (client identifies automatic thoughts but doesn’t believe alternatives at any meaningful level — believability consistently below 30/100)
  2. Behavioral component may be underdeveloped — “I do the activities but feel detached from them”
  3. Possible schema-level material maintaining partial response

Decision: shift focus from automatic thought work to behavioral experiments and introduce schema-level exploration. Review treatment plan together.

Risk: No SI. PHQ-9 item 9 = 0.

Homework: Pause thought records. One behavioral experiment per week: do something that goes against a rule you live by. Bring observations. Success criterion: one experiment with written observations.

Next session: Review behavioral experiment. Begin schema identification if appropriate.

Clinical commentary: A plateau documented honestly is more clinically useful than notes that minimize it. The declining homework completion is tracked numerically across sessions — this is what makes the pattern visible. The clinical reasoning for changing approach is written out explicitly.

Common CBT documentation mistakes

Automatic thoughts recorded as summaries, not quotes. “Client has catastrophic thinking about work” is not a CBT note. “Client’s automatic thought: ‘If I make one mistake, my manager will lose all confidence in me’” is.

Homework described without a success criterion. “Practice thought records” is not a homework assignment. “Complete one thought record for any situation rated ≥5/10 on distress” is.

Completion logged without clinical meaning. “Homework completed” tells you nothing. Did it work? What did the client learn? What does partial completion suggest about maintaining factors?

No scale scores. PHQ-9 and GAD-7 are 2-minute investments that give your Assessment section objective data. Without them, you’re estimating by impression.

Assessment that restates Subjective. “Client is anxious about work” is not an assessment. “Client’s anxiety about work is maintained by avoidance of the manager check-in, now three weeks running; this is the primary treatment target for next session” is.

TheraMemory keeps CBT session notes in a structured client timeline — scale scores tracked chronologically, homework visible from session to session, and your own preparation notes available before each appointment.

Try TheraMemory free for 14 days

Structured session notes with scale tracking, full client history, encrypted storage — built for the way CBT documentation actually works.

Document every session in under 10 minutes.

Try TheraMemory

Read also