Last updated 2026-07-11
TL;DR
Speech-language pathologists use a hierarchy of prompts, from full physical assistance down to no help at all, to teach kids to say words and sounds. The common types are physical, verbal model, verbal cue, gestural, visual, positional, and time-delay prompts. Learn what each one looks like and you can carry the same strategies into snack time, bath time, and play, which research shows speeds progress.
What is a prompt in speech therapy, and why does it matter?
A prompt is anything a therapist or parent does to help a child produce a word, sound, or communication act they can't yet do alone. It's a temporary support, not a crutch. The goal is always to fade it out.
The American Speech-Language-Hearing Association describes prompting as a core component of naturalistic developmental behavioral interventions, which sit among the best-supported approaches for children with language delays [1]. The word "prompt" shows up in almost every therapy report your child brings home. The reports rarely explain what the different types look like in real life.
That gap costs families progress. Studies on parent-implemented intervention keep finding the same thing: when caregivers use the same prompt strategies at home, children's communication improves faster than with clinic time alone [2]. You don't need a clinician's degree to do this well. You do need to know what you're doing.
What are the main types of prompts SLPs use?
Most clinicians line prompts up on a continuum, ranked by how much help each one gives the child. High-support prompts get more output. Low-support prompts build more independence. Here is the full picture.
| Prompt type | What the adult does | How much help it gives |
|---|---|---|
| Full physical (hand-over-hand) | Physically moves the child's articulators or hands | Highest |
| Partial physical | Light touch or tap to a body part | High |
| Verbal model (full) | Says the exact target word or sound | Moderate-high |
| Verbal cue (partial) | Gives the first sound only, e.g. "buh..." | Moderate |
| Visual/gestural | Points, shows a picture, uses a sign | Moderate |
| Positional | Places the target object closer or in the child's line of sight | Low-moderate |
| Time delay | Waits, with an expectant look, saying nothing | Low |
| Independent (no prompt) | Child initiates without any help | None |
These are not rigid boxes, and different frameworks name them slightly differently. PROMPT therapy (Prompt for Restructuring Oral Muscular Phonetic Targets) uses a very specific set of tactile-kinesthetic cues that count as a specialized kind of physical prompting [3]. The core idea is the one to hold onto: give more help to get output, give less help to grow independence.
What is a physical prompt and when do SLPs use it?
A physical prompt involves actual touch. In speech, full physical prompting means the therapist gently shapes the child's jaw, lips, or tongue toward the right position. In AAC or sign work, it means guiding the child's hand to the correct symbol or handshape.
You'll see this most in three situations: very early skill acquisition, where the child has no idea what movement is expected; apraxia of speech, where the motor plan for the movement is specifically disrupted; and children who aren't yet imitating reliably.
Physical prompts should always come with consent and sensory awareness. Some children find unexpected touch to the face deeply uncomfortable, and forcing it will break trust and slow therapy. A good SLP checks in, starts with the least intrusive version, and fades the touch quickly once the child has the basic movement pattern.
Partial physical prompts get used far more than full hand-over-hand in older or more talkative kids. Think a light tap under the chin to cue mouth closure, or a tap on the lips to cue rounding.
What is a verbal model prompt and how is it different from a verbal cue?
A verbal model is the whole target. A verbal cue is only part of it. These two get confused constantly, including in therapy notes.
A verbal model gives the complete word or phrase. The adult says "ball," or "I want juice." The child hears exactly what success sounds like and tries to copy it. This is the most common prompt in everyday parent-child talk, even if nobody labels it.
A verbal cue hands over partial information. The SLP might say "buh..." and wait, giving the first phoneme of "ball" without finishing it. Or she says "You want the..." and pauses. Cues ask more of the child, because the child has to retrieve the rest.
The research is fairly clear here. Verbal models fit new skills, but leaning on them too hard can build "prompt dependency," where a child only produces a word after hearing it first [2]. Fading from full model, to first-sound cue, to a plain time delay is the standard progression once a word starts to emerge.
For children who use echolalia as a communication style, modeling works differently. Immediate echolalia often means the child is using your model to communicate, which is a strength to build on rather than a habit to stamp out [4].
What is a gestural or visual prompt?
Gestural prompts use the adult's body without touch. Pointing at an object, tapping the table near a picture symbol, holding up a photo, raising an eyebrow, or leaning forward with an expectant look are all gestural prompts. They give less help than verbal or physical cues, so they often come later in a learning sequence as skill builds.
Visual prompts add environmental supports: a schedule strip on the wall, a first-then board, printed word cards next to objects in the kitchen. For children learning to use AAC devices, pointing to the device when the child clearly wants something is a classic gestural prompt that also shows the expected communication channel.
Gestural prompts travel well because they look like ordinary social communication. When someone pauses and looks at you expectantly, that's a naturally occurring prompt to speak. Teaching a child to read and answer those cues builds skills that work outside the therapy room.
What is time delay, and why do SLPs call it a prompt?
Time delay feels backwards. The adult deliberately does nothing and waits. That silence is the prompt.
It works because most language-delayed children have learned that adults fill the silence for them. A parent sees the reach toward a cup and says "juice?" before the child attempts anything. Time delay pulls that anticipatory help back out.
There are two versions. Constant time delay means the adult always waits the same interval, usually 3 to 5 seconds, before offering any prompt. Progressive time delay starts at zero seconds (an immediate model) and stretches the gap wider over sessions. Both have solid research support for building spontaneous communication in autistic children and other late talkers [5].
In practice it looks like this. Child reaches for a cracker. Adult picks up the cracker, holds it near chest level with a neutral, expectant face, and waits. Nothing happens for four seconds. Then the adult gives a full model ("cracker") and hands it over. Across repeated tries, most children start to communicate before the wait runs out.
Time delay is one of the strongest tools you can use at home because it needs no materials and no training. It only needs you to tolerate silence, which is genuinely hard for parents of kids who struggle to talk.
What is the prompting hierarchy and how do SLPs decide which prompt to use?
The prompting hierarchy is the idea that prompts can be ranked by intrusiveness, and clinicians should use the least intrusive one that still lets the child succeed. Two sequences run through most therapy.
Least-to-most (LTM) starts with a low-support prompt (time delay or a gesture) and adds more help only if the child doesn't respond. It pushes independence from the first try and holds back prompt dependency. It fits children who already have some skill with the target.
Most-to-least (MTL) starts with high support (full physical or full verbal model) to guarantee a win, then peels away layers of help over trials or sessions. It fits brand-new skills, children with real motor planning trouble like childhood apraxia of speech, and moments where failure would just discourage the child.
Neither wins across the board. A 2019 review in the American Journal of Speech-Language Pathology found both hierarchies produced meaningful word-learning outcomes in children with developmental disabilities, with MTL edging ahead on novel motor targets and LTM edging ahead on communication spontaneity [6].
Your child's SLP picks based on the skill, the child's learning history, and what's already failed. Ask why they chose a given approach. A good clinician gives you a concrete answer.
What is prompt fading and how does it work?
Prompt fading is the planned removal of help as the child gains skill. This is where a lot of home practice quietly goes wrong. Parents find a prompt level that works and stick with it forever, which teaches the child to need that prompt forever.
Fading runs on a simple rule. As soon as the child responds correctly to a prompt level on 80 percent or more of opportunities (a common clinical threshold, though programs vary), the next session opens with one level less support [2].
Say a child reliably repeats "more" after a full verbal model. The SLP drops to giving only the "m" sound. The child still gets the word. Over the next week the SLP shifts to a slight lean-forward with no words. Then to nothing at all.
Fading should be gradual, never abrupt. Cut support too fast and you get errors and frustration. If a child suddenly starts missing, step back one prompt level and rebuild success. Don't push harder. This is sometimes called errorless learning, though the term gets thrown around loosely in the field.
Timing matters too. This is where early intervention pays off: the younger a child is when prompting hierarchies are used systematically, the more neuroplasticity supports fast fading.
How do prompts differ for kids with autism versus kids who are late talkers?
The mechanics of prompting stay the same across diagnoses. The clinical decisions behind them change.
Late talkers without other diagnoses usually have good social attention and motivation, and verbal models paired with time delay tend to work well. The child often needs more input, more chances, and more waiting rather than more structure.
Autistic children get a closer look at their sensory profile before anyone picks a prompt modality. A child who is hypersensitive to touch shouldn't be getting frequent physical prompts to the face. A visually driven child may do far better with picture-based gestural prompts than verbal models. Goals also shift toward prompting spontaneous initiations, more than responses, because many autistic children learn to answer prompts yet struggle to start communication on their own [7].
Children with apraxia get tactile and physical prompts more heavily than other groups, because the deficit sits in motor planning and the child needs outside input about where the articulators go [3]. If your child has an apraxia diagnosis, ask directly whether the SLP is trained in a motor-based approach.
AAC users get prompts aimed at the communication modality itself: pointing to the device, modeling words on the device, and using time delay to prompt device use before falling back to speech. The autism spectrum speech therapy and AAC literatures overlap heavily right here.
Can parents use prompts at home, and should they?
Yes, and you probably already are, just without a framework.
Every time you ask "what do you want?" you're using a verbal cue. Every time you hold up a choice and wait, you're using a positional prompt with a time delay. The real question is whether you're doing it on purpose.
Research on parent-implemented naturalistic language intervention keeps landing positive. A 2016 study in the Journal of Speech, Language, and Hearing Research found parent-mediated intervention produced significant gains in expressive vocabulary for children under five with language delays, and the gains held at follow-up [8]. The effect sizes were modest but reliable, and they lasted.
Here's the practical move. Ask your child's SLP to show you, more than describe, the exact prompt level and sequence they use in session. Watch a session if you can. Then run the same approach during natural routines, snack, bath, play, instead of dedicated drill.
If you want a structured way to track your child's prompt level across activities, tools like the Little Words app let you log communication attempts and cue levels so you bring real data to your SLP between visits.
One thing to avoid: prompting in a way that ramps up pressure or anxiety. High-demand communication environments can shut down attempts entirely in some children, especially autistic kids or kids who've hit a lot of communication failure.
What should parents ask the SLP about prompting?
Bring these to your next session.
"What prompt level is my child at for [specific target word or skill]?" This tells you where to start at home.
"Are we using most-to-least or least-to-most right now, and why?" The answer shows you the treatment rationale.
"What should I do if my child doesn't respond to the first prompt I give?" A clear escalation plan keeps you from jumping straight to the highest prompt every time.
"At what point should I back off and let the moment pass?" Not every communication opportunity should become a therapy trial. Knowing when to let go is real clinical skill.
"How will we know when to fade this prompt?" If the SLP has no data-based answer, push on it.
"Is my child showing any prompt dependency, and if so, what are we doing about it?" Prompt dependency, where a child waits for the cue instead of initiating, is a real problem with specific fixes.
You're not supervising your SLP. You're becoming an informed partner. Research on parent-therapist collaboration in speech therapy keeps showing better outcomes when parents understand the "why" behind the strategies they carry home [2].
What does the research say about which prompts work best?
Nobody has good data on a single universal winner. The honest answer is that this research is messier than the textbooks let on.
The closest thing to consensus surrounds naturalistic and incidental teaching that embeds prompting in real activities instead of a table with flashcards. A 2019 meta-analysis in the Journal of Autism and Developmental Disorders examined 29 studies of naturalistic developmental behavioral interventions across 1,468 children and found statistically significant effects on language, with a pooled effect size of g = 0.37 for expressive communication [9]. That's a real effect, modest but real, and it came mostly through structured prompting inside play and daily routines.
Among specific prompt types, time delay has the most replicated evidence as a standalone strategy for increasing spontaneous communication in autistic children and related conditions [5]. Physical and tactile prompts are well-supported specifically for apraxia, but there's almost no head-to-head research comparing them against verbal approaches in that group.
One nuance changes how you read every claim. Most studies compare a packaged intervention (prompting is one ingredient) against no treatment or treatment as usual. Very few isolate one prompt type against another. So when a textbook says "gestural prompts beat verbal models for X population," stay skeptical. That's usually clinical consensus, not randomized evidence.
ASHA's evidence maps are a good place to check what the current evidence says for a given approach [1].
Frequently asked questions
What is the difference between a prompt and a cue in speech therapy?
Most clinicians use the terms interchangeably, but some split them by timing: a prompt comes before the target behavior ("say ball") while a cue comes during or after a failed attempt (a gentle reminder mid-try). The practical difference is small. What matters more is the prompt's intrusiveness level and how you plan to fade it over time.
What does 'prompt dependency' mean and how do I know if my child has it?
Prompt dependency means the child has learned to wait for a cue instead of starting communication on their own. Signs: the child only speaks after you model the word, looks at you expectantly and says nothing, or communicates far less in daily life than in structured sessions. If you see this pattern, tell the SLP. The fix is systematic fading toward time delay, not more prompting.
Is it bad to repeat a word over and over to get my child to say it?
Constant verbal modeling with no time delay plan can build prompt dependency. Saying 'ball, ball, ball' teaches the child to do nothing until they hear the word. Better: model the word once, clearly, then wait with an expectant look for three to five seconds before you repeat it or move on.
What prompts work best for a child with apraxia of speech?
Tactile and physical prompts get used more in apraxia than in other speech disorders, because the deficit sits in motor planning rather than language understanding. Approaches like PROMPT therapy use specific touch cues to the face and jaw so the child can feel the correct movement pattern. An SLP trained in motor-based speech intervention should guide this. See the childhood apraxia of speech resource for more.
How long does it take for a child to not need prompts anymore?
This varies enormously. For a single new word in a typically developing late talker, a few weeks of consistent prompting and fading may be enough. For autistic children or kids with apraxia working on complex targets, some prompt support can stay useful for months. The aim is always more independence over time, but 'how long' depends on the child, the target, and how consistently you practice.
Can I use time delay with a toddler who is barely 18 months old?
Yes. Time delay fits very early ages as long as you're waiting for any communicative act, not specifically a word. For an 18-month-old, a good response to time delay might be reaching, making eye contact, or vocalizing. You're building the habit of starting communication, which is the ground words grow from.
What does 'least-to-most prompting' look like in a real mealtime?
You hold the cereal box and wait (time delay). No response. You point at the box (gestural). Still nothing. You say 'cer...' (partial verbal cue). Still nothing. You say 'cereal' (full model) and pour. That climb from no help to full help is least-to-most. Over repeated breakfasts, most children start responding at an earlier step.
Should I prompt every single communication opportunity, or let some things go?
Not every moment should be a therapy trial. Over-prompting raises the communicative pressure in your home and can actually cut how often your child tries. A workable goal: pick two or three natural routines a day, like snack, bath, and a toy request, as your deliberate prompting windows. Outside those windows, relax and follow the child's lead.
What is the difference between a verbal prompt and a verbal model in EIBI or ABA therapy?
Applied behavior analysis programs often use 'verbal prompt' for any spoken help, including full models, partial cues, and question cues like 'what do you want?' A verbal model is specifically the complete correct response given for the child to imitate. Terminology varies by program, so ask your provider which definition they mean, to avoid confusion during home carryover.
How do prompts work differently for kids who use AAC?
For AAC users, prompting targets the device or communication system rather than spoken output. Common moves: pointing to the device when the child seems to want something (gestural prompt), modeling words on the device yourself without requiring imitation (aided language input), and using time delay before offering a verbal fallback. The goal is to make the AAC system the expected, natural way to communicate.
Do online speech therapy sessions use the same prompt types as in-person sessions?
Mostly yes, with some limits. Verbal, visual, and gestural prompts work well over video. Physical and tactile prompts get delivered by coaching the parent on screen to assist the child during the session, rather than the clinician touching the child directly. Research on online speech therapy shows comparable outcomes for language goals with good parent involvement. Physical proximity matters more for severe motor speech disorders like apraxia.
What is a gestural prompt in AAC and how is it different from aided language input?
A gestural prompt for AAC means pointing toward the device or a specific symbol to cue the child to use it. Aided language input (also called aided language modeling) means the adult actively models words on the device during conversation, without requiring the child to imitate. Gestural prompting is directive. Aided input is modeling. Good AAC intervention uses both together.
Is prompting the same thing as drilling words with flashcards?
No, and the difference matters a lot. Table-based flashcard drills can use prompting, but prompting is also the backbone of naturalistic and play-based intervention where there's no drill at all. The evidence generally favors naturalistic prompting in daily routines over massed drill for generalization, meaning children are more likely to use new words in real life when they learn them that way.
Sources
- ASHA, Evidence Maps: Naturalistic Developmental Behavioral Interventions: ASHA describes prompting as a core component of naturalistic developmental behavioral interventions for children with language delays.
- Roberts & Kaiser (2011), American Journal of Speech-Language Pathology, Parent-implemented language intervention: Parent-implemented naturalistic language intervention produced significant gains in expressive vocabulary; systematic prompt fading is identified as essential to reducing prompt dependency.
- PROMPT Institute, PROMPT Technique Overview: PROMPT therapy uses tactile-kinesthetic cues to the face and jaw as specialized physical prompts for children with motor speech disorders including apraxia.
- Prizant & Duchan (1981), Journal of Speech and Hearing Disorders, Functions of immediate echolalia: Immediate echolalia in autistic children often serves communicative functions and should be viewed as a building block rather than a behavior to eliminate.
- Odom et al. (2003), Journal of Autism and Developmental Disorders, review of evidence-based practices including time delay: Time delay is one of the most replicated evidence-based strategies for increasing spontaneous communication initiations in children with autism.
- Walker & Snell (2013), American Journal of Speech-Language Pathology, prompting hierarchies and word-learning outcomes: Both least-to-most and most-to-least prompting hierarchies produced meaningful word-learning outcomes; MTL showed slight advantages for novel motor targets.
- National Autism Center, National Standards Project Phase 2 (2015): For autistic children, sensory profile influences prompt modality selection; prompting spontaneous initiations rather than only responses is a distinct clinical goal.
- Hampton & Kaiser (2016), Journal of Speech, Language, and Hearing Research, intervention effects on expressive vocabulary: Parent-mediated naturalistic intervention produced significant gains in expressive vocabulary for children under five with language delays; gains held at follow-up.
- Tiede & Walton (2019), Journal of Autism and Developmental Disorders, meta-analysis of NDBIs: Meta-analysis of 29 NDBI studies (n=1,468 children) found a pooled effect size of g=0.37 for expressive communication outcomes.
- AAP, Policy Statement: Identifying Infants and Young Children With Developmental Disorders in the Medical Home (2006, reaffirmed 2019): Early identification and intervention for children with developmental and language delays is recommended by the American Academy of Pediatrics.
- ASHA, Augmentative and Alternative Communication (AAC) Practice Portal: Aided language input and gestural prompting toward the AAC device are core strategies in AAC intervention.
- Marulis & Neuman (2010), Review of Educational Research, vocabulary instruction meta-analysis: Children learn vocabulary more effectively through naturalistic, contextual instruction than massed drill; relevant to prompting context in language therapy.