A SOAP Note Template Built for the Speech Session You Just Ran

You have five minutes before the next client, and the note for the last one is still open. A generic SOAP template does not help much in that gap, because its prompts are written for any clinician anywhere: “objective findings,” “assessment,” “plan.” What a speech-language session actually produces is more specific than that — a target, practiced at a level, in a task, under a cue level, at a measured accuracy. A template that asks for exactly those things gets filled in faster and produces a note that means something three weeks later. This article gives you that template, then works it through an articulation session, a language session, and a fluency session so you can see what changes and what does not.

Strip any defensible speech treatment note to its skeleton and the same four facts appear for every goal addressed: which target (the /r/ sound, regular past tense, easy onset), at which level (isolation, word, phrase, sentence, conversation), under which cues (modeling, visual prompts, phonemic cues, and how often you gave them), and at what accuracy (correct trials over total trials). Those four facts are what make the next session plannable and the episode reviewable. A note that says “worked on articulation with good progress” contains none of them, and a note that contains all four barely needs anything else in the objective section.

The cue level is the one clinicians most often drop, and it is the one that carries the skilled-service argument. Accuracy alone is ambiguous: 80% with continuous models is a different clinical picture from 80% independent, and only the first one explains what the therapist was doing. ASHA’s documentation guidance frames the objective section as recording both the patient’s performance and the skilled intervention you provided in response — the cue line is where those two meet.

SOAP is a convention, not a regulation. What Medicare actually requires of a treatment note is a short list, set out in the Medicare Benefit Policy Manual (Pub 100-02, Chapter 15, Section 220.3): the date of treatment, an identification of each specific intervention or modality provided and billed, the total treatment time, and the signature and professional identification of the qualified professional who furnished the services. For timed codes, the total timed-code minutes must be documented too — though most SLP treatment codes are untimed, so for many speech sessions the single total-time line does the work. The template below maps every one of these to a labeled line, so completing it honestly is the compliance work.

That 2027 change is worth a habit adjustment now. Because 92507 is untimed, many SLPs never built the reflex of recording session minutes — there was nothing to add up. The replacement codes are structured like the timed codes physical and occupational therapists already live with: a base code for the first 30 minutes of direct one-on-one treatment and an add-on for each additional 15. When your payers adopt them, the minutes line in your note stops being a formality and starts being the thing the claim is built from. Recording start, end, and total time costs one line today and means the transition changes nothing about how you write. For how minute-counting works once codes are timed, see timed vs. untimed codes.

The template repeats one block — the target block — once per goal addressed in the session. That repetition is the design: a speech session is usually two to four targets deep, and the note should show each one’s data separately rather than averaging them into a paragraph. Everything outside the target blocks stays short. If your subjective line is regularly longer than your objective data, the note is upside down.

All three examples below are fictional composites written for illustration. What is worth noticing is how little changes between them: the target block’s four facts stay the same, and only what counts as a “trial” and what counts as a “cue” shifts with the domain.

Read the three objective sections side by side and the pattern is the same sentence with different nouns: task, count, cue. The articulation trial is a word, the language trial is an obligatory context, the fluency “trial” is a sentence-initial opportunity to use the technique — but each one is countable, each is tied to a stated cue level, and each produces a number the next session can be planned from and the next note can be compared against.

There is no single mandated cue taxonomy, and you do not need one. What you need is for the words in your cue line to mean the same thing in every note you write, because the fading of cues across sessions is often the clearest evidence of progress a speech chart contains — but only if the labels are stable enough to compare. A workable shorthand has two axes: what kind of cue, and how often you gave it.

1. 01

   Name the cue type in the client’s terms

   Model (you produce it first), phonemic or placement cue (“lips round”), visual (gesture, mirror, written prompt), sentence-completion or cloze, expectant pause. Pick the four or five you actually use and reuse those exact words.

2. 02

   State the frequency, not just the type

   Continuous (every trial), intermittent (after errors, or every nth trial), faded (started continuous, withdrawn mid-task), none. “Minimal cues” alone is a judgment; “intermittent visual cue after errors” is a condition someone else could reproduce.

3. 03

   Record the direction of change

   When today’s cue level differs from last session’s, say so in the cue line itself — “faded from continuous model” — so the trend is readable inside a single note, not only by pulling the whole chart.