The hardest part of clinical documentation isn't choosing a format — it's knowing what "good" looks like inside the format. This page shows side-by-side examples of weak versus strong documentation, organized by the documentation best practice each pair illustrates.
Each example pair is short — these aren't full session notes — but they capture the precise difference between documentation that passes audit and supervision review and documentation that doesn't. For a deeper look at the four note formats themselves, see our therapy note formats guide.
Start Free TrialEach section below pairs a weak version of a documentation snippet with a stronger version, followed by a brief explanation of why the stronger version works. Both versions describe the same clinical situation — the only thing that changes is the specificity and structure.
These are reference models, not phrasing to copy. Strong documentation reflects your actual session content; copying language verbatim across notes is a common audit flag.
The single highest-leverage documentation habit is naming the specific technique used, not just the modality. "Used CBT" tells a reader almost nothing. The strong version names the technique, what it targeted, and why it was clinically appropriate.
Used CBT to address client's anxiety.
Used Socratic questioning to test client's catastrophizing thought about Friday's work review (e.g., 'I'll definitely be fired'). Walked through evidence-for/evidence-against worksheet in session.
Why it works: The strong version names a specific CBT technique (Socratic questioning), the specific cognitive distortion targeted (catastrophizing), the clinical context (work review), and the in-session activity (worksheet). It demonstrates clinical thinking at the technique level.
Worked on emotion regulation.
Practiced TIP (Temperature, Intense exercise, Paced breathing, Paired muscle relaxation) skills from DBT distress tolerance module. Client identified two situations from the past week where she could have used TIP rather than self-harming.
Why it works: Names the DBT module (distress tolerance) and the specific skill (TIP), and ties skill use to a concrete situation in the client's life. A reader can see exactly what was practiced and how the client engaged with it.
Insurance reviewers and supervisors look for measurable progress — something more concrete than "client is improving." Standardized screener scores, behavior frequency counts, and trend language all qualify. Strong progress documentation is specific and comparable across sessions.
Client is making good progress.
GAD-7 score 11 this session, down from 15 at intake four weeks ago. Client also reports anxiety as 5/10 average daily (down from 7/10 last week) and used breathing exercise before stressful work meeting on Wednesday.
Why it works: Combines a standardized measure (GAD-7), a self-report rating, and a concrete behavioral data point. A reviewer can see the trend, compare across sessions, and verify that progress is real rather than impressionistic.
Checking behavior is decreasing.
Client reports checking-behavior frequency decreased from 8x/day at baseline to 3x/day this week. Independently using cognitive restructuring 4 of 7 days. Goal is to reduce checking to under 2x/day by week 12.
Why it works: Quantifies the baseline, the current level, and the goal. Includes a related behavioral marker (cognitive restructuring use). Makes treatment effectiveness visible rather than asserted.
A reader should be able to trace a clear line from any session's content back to the treatment plan. Sessions that don't connect to named treatment goals raise medical-necessity questions. The strong version makes the connection explicit.
Discussed client's relationship issues.
Session focused on treatment plan goal #2 (improve boundary-setting in close relationships). Client described a recurring pattern with her mother where she agrees to obligations she resents. Worked on identifying the moment of agreement and the somatic cues that precede it.
Why it works: Names the treatment plan goal explicitly, ties session content to that goal, and frames the work clinically. The note now demonstrates medical necessity and shows continuity of care.
Observation is what you saw. Interpretation is what you concluded. Mixing them — especially in the Objective section of a SOAP note or the Behavior section of a BIRP note — weakens the chart's defensibility. Keep observable facts and clinical inference separate.
Client was anxious and depressed today.
Client reported feeling 'really anxious' and 'low.' Observed: pacing during the first five minutes of session, intermittent eye contact, rapid speech, and tearfulness when describing the past week. PHQ-9 score 14, GAD-7 score 13.
Why it works: Separates what the client said (Subjective: 'really anxious,' 'low') from what was observed (Objective: pacing, eye contact, speech rate, tearfulness, scores). The clinical interpretation — that mood is dysregulated — belongs in Assessment, not in either of these sections.
"Continue therapy" is not a plan. The Plan section should give a reader a concrete answer to: what specifically happens next, what homework was assigned, and when the next contact is.
Continue therapy. Follow up next week.
Continue exposure-based work targeting uncertainty tolerance. Assigned thought record for next week, focused on health-related automatic thoughts. Planning to introduce behavioral experiment in next session if anxiety continues trending down. Re-administer GAD-7 in two weeks.
Why it works: Names the next clinical move (continued exposure work), the specific homework (thought record), the contingency for the following session (behavioral experiment if trend holds), and the measurement plan (GAD-7 in two weeks). A reviewer can see the treatment trajectory.
Every clinical note should at least briefly address risk when it's clinically indicated — and many auditors expect a one-line statement even when risk is low. Silence on the topic is the most common documentation gap that creates liability exposure.
No safety concerns.
Denied SI/HI; no plan or means; protective factors include strong therapeutic alliance, custody of two children, and active engagement in treatment. Safety plan from intake reviewed and remains current.
Why it works: The strong version names what was assessed (SI/HI), what was ruled out (plan, means), what protects the client (specific protective factors), and the status of any existing safety plan. It's still brief — three lines — but documents the clinical thinking.
Beyond the section-by-section examples above, these are the documentation patterns that most often surface during audits and supervision review:
Copy-pasting the same paragraph across multiple sessions — strong indicator of templated rather than session-specific documentation, and a common audit flag.
Writing notes more than 24 hours after the session — memory degrades and notes become less accurate, less specific, and harder to write.
Letting note length drift up over time — most reviewable notes are 150–400 words; longer is rarely better.
Using diagnostic shorthand without naming the working impression — e.g., writing 'GAD' without ever stating that as the working diagnosis.
Documenting plans in passive voice — 'will be addressed' tells a reader less than 'will introduce thought records targeting health-related automatic thoughts.'
Skipping risk documentation when it's clinically indicated — even one line ('denied SI/HI; no acute concerns') is dramatically better than silence.
Writing 'progress noted' without quantifying it — without a measurable, the chart can't demonstrate that treatment is working.
Good documentation is concise, clinically specific, and easy for any reviewer — supervisor, insurer, or future provider — to follow. Strong notes name interventions specifically, document measurable progress, connect each session to treatment goals, separate observation from interpretation, and include actionable plans. Length is not the criterion — most well-written notes are 150–400 words.
No. Use them as reference models — every note should reflect your actual session content. Copy-pasting documentation across notes is a common audit flag. The examples here demonstrate the level of specificity and structure that strong documentation should have, not literal phrasing to reuse.
Most reviewable therapy notes land between 150 and 400 words. Length should match clinical complexity — routine sessions can be brief, while intake, high-risk, or significant clinical change sessions warrant more detail. Longer notes are not better. Overly long notes increase documentation burden without improving clinical care.
These examples demonstrate the documentation features insurance reviewers look for: medical necessity (named diagnosis or working impression), specific interventions (technique-level, not modality-level), and measurable progress against treatment goals. Reviewers don't dictate format, but they do expect this level of clinical specificity regardless of whether you use SOAP, DAP, BIRP, or progress-note structure.
The principles apply universally. CBT, DBT, psychodynamic, EMDR, family systems, and other modalities all benefit from naming specific interventions, tracking measurable progress, and connecting sessions to treatment goals. The specific intervention names will differ by modality, but the documentation principles are the same.
Same-day documentation, even when brief, is dramatically better than delayed documentation. Memory degrades fast — a 60-second note written between sessions is more accurate than a longer note written three days later. Pre-built templates and AI-assisted drafting can reduce documentation time without sacrificing the clinical specificity these examples demonstrate.
AfterSession produces well-structured therapy notes from a brief session summary — naming specific interventions, tracking measurable progress, and connecting to treatment goals.