Guided meditation script template and pre-recording checklist

Teachers who improvise well in a live room often assume a script will make them sound stiff. In practice the opposite is usually true on a recording, because a script removes the cognitive load of deciding what comes next — and that load is audible. It shows up as hesitation, filler words, sentences that restart, and a subtle tension in the voice.

A script also lets you control the two things that matter most in recorded meditation: the length of each segment and the placement of silence. Improvised sessions tend to be front-heavy, with a long settling section and a rushed close.

The goal is not to read. It is to know the shape well enough that you can speak it. Write the script, read it aloud twice, then record it while looking at it rather than reading from it word by word.

Almost every effective guided meditation, regardless of tradition or length, moves through the same five stages. The proportions shift; the sequence rarely does.

Arrival, roughly ten per cent. You orient the listener: what this session is, how long it will take, what position to be in. Short and practical. This is also where the listener decides whether to trust your voice for the next twenty minutes.

Settling, roughly twenty per cent. Posture, a few breaths, the transition from doing to attending. Resist the urge to make this the whole session; its job is to hand over to the practice.

Body or anchor, roughly fifteen per cent. Establish the thing attention will return to — breath, body sensation, sound, a phrase. Name it clearly, because everything after this refers back to it.

Core practice, roughly forty per cent. The substance of the session, and the part with the longest silences. Instruction density should be at its lowest here.

Return, roughly fifteen per cent. Widen attention back out, reorient to the room, and close. Never end abruptly from deep practice — the transition out is part of the practice, not an afterthought.

Written prose and spoken guidance follow different rules. A sentence that reads elegantly can be almost unfollowable with the eyes closed, because a listener cannot go back.

Keep sentences short — one instruction each. Long subordinate clauses force the listener to hold a structure in working memory, which is precisely the mental activity you are asking them to set down.

Stay in the present tense and use permissive rather than commanding language. “You might notice” and “if it feels comfortable” leave room for a listener whose experience does not match your script. “Feel your shoulders relax” fails the moment their shoulders do not.

Watch for hard consonant clusters and plosive-heavy phrases when you read aloud. Meditation vocabulary is full of p and b sounds and they land in your quietest passages. If a phrase pops, rewrite it rather than fighting it at the microphone.

Repeat deliberately. Returning to the same phrasing for the same instruction gives a session an internal rhythm and reduces the listener's cognitive load. On the page it looks repetitive; in the ear it feels like structure.

Silence is the part of the script most likely to be lost in the moment. Write it in explicitly, with a duration, and treat it as a line rather than as a gap between lines.

A simple notation is enough. Use a bracketed marker on its own line: [pause 3s], [pause 8s], [pause 30s]. Making it visually distinct means you see it coming rather than reading past it.

Reasonable defaults: two to three seconds after a simple instruction, five to eight seconds after inviting someone to notice something, fifteen to thirty seconds during extended practice, and three to five seconds of quiet after your final word before the file ends.

Record the silences live rather than inserting them afterwards. Live pauses carry your room tone underneath them; inserted digital silence creates an audible hole in the ambience that is more noticeable than the pause itself.

What follows is a skeleton for a ten-minute session. Scale the pauses and the core section for longer formats; the sequence stays the same.

Arrival — about one minute

“Welcome. This is a [length] practice for [purpose]. You can sit or lie down — whatever lets you be still and comfortable. [pause 3s] When you are ready, let your eyes close, or let your gaze rest softly somewhere in front of you. [pause 5s]”

Settling — about two minutes

“Let the weight of your body be held by whatever is beneath you. [pause 5s] Take one slower breath in. [pause 3s] And let it go. [pause 5s] There is nothing to achieve in the next [length]. You are just here. [pause 8s]”

Anchor — about ninety seconds

“Bring your attention to [anchor — the breath, the hands, the sounds in the room]. [pause 5s] You do not need to change it. Just notice it. [pause 8s] Whenever you notice you have drifted, that noticing is the practice. Come back, gently. [pause 8s]”

Core practice — about four minutes

“[Core instruction — one or two sentences.] [pause 30s] [Light reminder — one short phrase.] [pause 30s] [Reminder.] [pause 30s]” — note how little you say here. Speak roughly every twenty to thirty seconds, and say less each time.

Return — about ninety seconds

“Let your attention widen out from [anchor] to your whole body. [pause 8s] Notice the room around you — the sounds, the light, the air. [pause 8s] When you are ready, and in your own time, let your eyes open. [pause 5s] Thank you for practising. [pause 5s]”

Almost every re-record is caused by something on this list. It takes about five minutes and it will save you a session.

Record the session straight through rather than section by section. Recording in pieces produces small mismatches in tone, pace and level at every join, and those joins are audible in a quiet mix.

If you stumble, do not stop. Pause for a couple of seconds, go back to the beginning of the sentence and say it again cleanly. The pause makes the mistake easy to find and the clean retake easy to keep.

Take the pauses in real time, at full length. They will feel much longer to you than they will to a listener with their eyes closed, and cutting them short in the moment is the single most common regret after listening back.

Because you recorded from a script, the transcript of your session is almost identical to the document you wrote. That makes text-based editing an efficient way to clean up: find the duplicated sentence where you restarted, delete the first version, and the audio follows.

Work through the transcript for the retakes first, then listen to the whole file once for the things text will not show you — clicks, plosives, level drift and pauses that turned out too short.

Then clean up noise lightly, add a music bed if you are using one, check the ending fades rather than stops, and normalise to the same loudness target you use for the rest of the series.