Writing notes

SOAP vs DAP vs BIRP: which note format should you actually use?

A practical comparison of the three formats you are most likely to meet, with a full worked example of each — written at the length you would actually write.

About 12 minutes · A guide to structuring a written record, not clinical advice

You finish a session. You open your template. And then you sit there, cursor blinking next to a heading that says "Assessment", with no clear sense of how much to write, how clinical to sound, or where your observation ends and your interpretation begins.

Almost nobody is taught note-writing properly. Most clinicians learn it by copying whoever supervised them first, then quietly wonder for years whether they are doing it wrong.

This article compares the three formats you are most likely to meet — SOAP, DAP and BIRP — and shows a complete worked example of each, written at the length you would actually write. By the end you should be able to choose one deliberately instead of inheriting one by accident.

A note on what this is: a guide to structuring a written record. It is not clinical guidance and not legal advice. What your notes must contain is set by your state law, your licensing board, your professional association, any payer contract you hold, and your employer's policy. Where a payer or program prescribes a format, theirs overrides everything below.

The short answer

If you want to stop reading after one paragraph:

• SOAP if your notes may be read by a prescriber, physician, or multidisciplinary team. It keeps the client's account and your observation in separate boxes, which makes it legible to people outside the therapy.
• DAP if you are in private practice and want the shortest defensible note. It merges those two boxes into one, because in a talking session the distinction is often artificial.
• BIRP if you work against a treatment plan, or with a funder or program that wants to see what you did and how the client responded. It gives your intervention its own section.

All three are equally acceptable in most settings. The choice is organizational, not clinical.

What all three have in common

Before the differences, the part that matters more: every one of these formats is a container for the same four kinds of information.

1. What happened — what the client reported, what you observed, what you did.
2. What you make of it — your clinical thinking, kept visibly separate from the facts.
3. Where it is going — the plan, the next appointment, anything agreed.
4. Risk — screened, and documented whether or not anything was found.

A note that contains those four things is a good note in any format. A note missing the fourth is the one that causes problems later, regardless of how neatly the rest is laid out.

Side by side

SOAP

Subjective · Objective · Assessment · Plan

SOAP came out of medical charting, and its logic is a physician's: keep what the patient tells you apart from what you can observe or measure. In therapy that separation is less natural than in medicine — but it is exactly what makes a SOAP note readable to someone who was not in the room.

When it earns its length

• The client is also under psychiatric or medical care, and your note may be shared with their authorization.
• You work in a clinic, day program, or any multidisciplinary team.
• You use screening measures routinely and want scores sitting in a predictable place.

A worked example

Fictional, written to show length and register.

Client: MR-0214 · Date: 12 March · Time: 09:00–09:50 · Session: 7 · Individual, in person

SSubjective

Client reports three episodes in the past ten days of sudden intense anxiety, twice while driving and once in a supermarket queue, describing "my chest closing up and being certain I'm about to pass out." Has begun taking a longer route to work to avoid the highway. Reports sleep unchanged, appetite reduced. Used the paced breathing agreed last session during two of the three episodes and found it "helped a bit, maybe took the edge off." Denies thoughts of self-harm.

OObjective

On time, engaged throughout. Speech at normal rate; some acceleration when describing the supermarket episode. Affect anxious, congruent with content. No psychomotor agitation observed in session. Anxiety measure score 14, up from 11 four weeks ago. Between-session task partially completed.

AAssessment

Panic symptoms with early avoidance behavior beginning to consolidate around driving and enclosed public spaces. The rise in the measure is consistent with the client's own account and appears to reflect increased frequency rather than increased severity. Alliance is strong and the client remains an active participant in planning. Main maintaining factor is safety behavior — route avoidance — which is currently reducing distress in the short term at the cost of reinforcing the pattern. Risk to self screened, none identified.

PPlan

Next session in seven days, in person. Focus: psychoeducation on the maintaining role of avoidance, and construction of a graded hierarchy beginning with a short highway drive at a low-traffic time. Client to record episodes with time, setting, and duration. Repeat the anxiety measure in four weeks. Consider liaison with prescriber if frequency increases; discussed with client, no consent sought at this stage.

What to watch for

The most common SOAP failure is putting interpretation in the Objective section. "Client was defensive" is an inference. "Client declined to discuss the argument and changed the subject twice" is an observation. Keep the second kind in O and move the first kind to A, where it belongs and where it can be reasoned about.

DAP

Data · Assessment · Plan

DAP is SOAP with the first two sections merged. That is the whole difference. It exists because for a great many therapy sessions, separating what the client said from what you noticed produces two paragraphs that repeat each other.

When it earns its brevity

• Independent private practice, where you are the only person who will read the note.
• Sessions whose substance is conversation rather than measurement.
• A full caseload, where saving four minutes per note saves you an hour a week.

A worked example

Fictional.

Client: JL-0071 · Date: 13 March · Time: 14:00–14:50 · Session: 11 · Individual, telehealth

DData

Session focused on the anniversary of the client's mother's death, which falls next week. Client had not raised the date in previous sessions and mentioned it in passing before returning to work matters; I named the omission and she agreed to stay with it. Described feeling "flat rather than sad" and being unsettled by her own lack of tears. Longer pauses than usual, and twice said "this is stupid" before continuing. We worked with the meaning she attaches to visible grief, drawing on her account of a family in which composure was expected. She has arranged to visit the grave with her brother, the first time they will have done so together. Reports sleeping better than last month; has resumed swimming twice weekly.

AAssessment

Anticipatory grief around a first anniversary, complicated by a family script in which the absence of overt emotion is read as failure to care. Her self-criticism about "not feeling enough" appears to be doing more distress than the grief itself. Capacity to stay with difficult material has clearly increased since the early sessions — she recovered from two urges to deflect without needing prompting. Resources: the relationship with her brother, and a returning exercise routine. No risk indicators; not formally screened this session as no concerns were present or reported.

PPlan

Next session in seven days, deliberately timed for two days after the anniversary. No between-session task set; agreed that adding a task this week would work against the point. If she finds the visit difficult she will email to bring the session forward. To raise in consultation: whether to invite the brother to a joint session later in the work.

What to watch for

Because DAP has no dedicated section for what you did, interventions quietly disappear. A note that records only what the client said and how you interpreted it does not evidence that therapy took place. Name your interventions inside the Data section — the example above says "I named the omission" and "we worked with the meaning she attaches" for exactly this reason.

BIRP

Behavior · Intervention · Response · Plan

BIRP restructures the note as a sequence: here is what presented, here is what I did about it, here is how the client responded, here is what happens next. It is the format most often required where somebody else is paying, because it answers the question a funder actually asks — what did you do, and did it work?

When it earns its length

• Work conducted against a treatment plan with measurable objectives.
• Insurance-funded or publicly funded programs, and anywhere subject to utilization review or audit.
• Addiction services, community mental health, and intensive outpatient settings.
• Supervision and training, because the intervention-to-response link is visible on the page.

A worked example

Fictional.

Client: DP-0139 · Date: 14 March · Time: 11:00–11:50 · Session: 5 · Individual, in person · Treatment plan goal 2, objective 2.1

BBehavior

Client reports two incidents in the past week of shouting during disagreements with his partner, one ending with him leaving the house for several hours. Rates his ability to notice rising anger before it escalates at 3 out of 10. Presented settled and self-critical, describing himself as "the problem in the relationship." Treatment plan objective 2.1 is to identify physical early warning signs of anger in at least three situations per week; the client's log shows one entry.

IIntervention

Chain analysis of the more serious incident, working backwards from leaving the house to the first bodily cue. Psychoeducation on the physiology of the anger escalation curve and why intervening early requires less effort than intervening late. Introduced a two-column monitoring sheet separating bodily cue from thought. Reframed the incomplete log as information about when monitoring is hard rather than as failure. Fifty minutes, individual, in person.

RResponse

Client identified jaw tension and a change in his breathing as reliable early cues, and located both roughly two minutes before the shouting began — earlier than he had believed possible. Visibly more animated during the chain analysis than in previous sessions. Accepted the reframe of the incomplete log without argument, which is a change; in sessions 2 and 3 similar reframes were met with further self-criticism. Agreed to the two-column sheet and set his own target of four entries. Rerated his ability to notice early cues at 5 out of 10 by the end of the session.

PPlan

Next session in seven days. Client to complete four monitoring entries. Session to review entries and introduce a single de-escalation strategy tied to the cues identified. Progress against objective 2.1: partial. Treatment plan unchanged. Next formal plan review in six weeks.

What to watch for

BIRP tempts you into writing a mini-essay. The Response section is the one that matters most and the one most often reduced to "client was receptive." Receptive to what, and how did you know? The example above gives a rerating and a behavioral change across sessions, both of which a reviewer can actually evaluate.

How to choose

Five questions, in order of how much they matter.

1. Does anyone else prescribe a format? A payer, program, agency, or employer. If yes, stop here and use theirs.

2. Might these notes be read by a prescriber or physician? If yes, SOAP. The separation it enforces is precisely what an outside reader needs.

3. Do you work against written, measurable treatment goals? If yes, BIRP. It is the only one of the three that makes progress against an objective visible without extra effort.

4. Are you the only person who will read these notes, and is time your constraint? If yes, DAP.

5. Still undecided? Use DAP and be disciplined about naming your interventions inside the Data section. It is the least effortful format to maintain, and consistency matters more than the choice.

Six things that make a note hard to defend

Format is the smaller half of the problem. These apply to all three.

Inference written as fact. "Client is manipulative" is an opinion in the voice of an observation. Describe the behavior; put your reading of it in the Assessment.

No record of what you did. A note describing only the client's material does not evidence a therapeutic intervention.

Risk not documented when nothing was found. "Risk to self screened, none identified" takes four seconds and is the entry you will be relieved to find later.

Third parties identified unnecessarily. Name the relationship, not the person, unless identifying them serves the record.

Notes written days later. Accuracy decays fast. Write the note the same day if you can, and if you cannot, note when it was written.

Silent corrections. Amend openly and date the amendment. Never overwrite the original entry — a record that appears altered is worse than one that was wrong.

Can you switch formats mid-treatment?

Yes, and it is unremarkable. Note the change and the date, and keep the new format consistently from that point. What causes difficulty is drift — three sessions in SOAP, two in something improvised, one that is a paragraph with no structure at all. A reviewer reading that sees a clinician who was not paying attention.

Frequently asked

Are psychotherapy process notes the same as the clinical record? In several jurisdictions, no — they receive additional protection and should be stored separately. All three formats above are for the clinical record. If you keep process notes, keep them apart, and check how your own jurisdiction treats them.

How long should a session note be? Long enough to show what happened, what you did, what you concluded, and what comes next. In practice that is usually 150 to 300 words. Consistently shorter than that and you may be leaving out your own interventions; consistently longer and you are probably writing a process note by accident.

Is one format more defensible than another? No. A well-written note in any of the three is defensible, and a vague note in any of the three is not. Specificity, contemporaneity, and the separation of observation from inference matter far more than the headings.

Which one is fastest? DAP, comfortably. Then SOAP. BIRP takes the longest and buys you the most when someone is reviewing your work.

In closing

The three formats are not competing philosophies. They are three arrangements of the same content, each optimized for a different reader — a colleague, yourself, or a reviewer. Work out who your reader is, choose accordingly, and then stop thinking about it, because the format is not the part that makes a note good.

What makes a note good is that it says what happened, distinguishes what you saw from what you concluded, records what you did, and can be read in two years by someone who was not there.

---

If you would like these three formats as ready-made templates — with a fully worked example of each, blank fillable versions, and an appendix on recording language and data minimization — I make a set of them here. They come as fillable PDF and editable Word, in four page sizes.

This article is about structuring written records. It is not clinical, legal, or billing advice, and it does not replace your training, your supervision, or the requirements of your jurisdiction. Where a payer, program, or employer prescribes a documentation format, follow theirs.