Every ABA practitioner knows the goal isn’t compliance — it’s independence. But the path from “requires full physical guidance” to “performs independently” depends entirely on how you deliver and fade prompts along the way. Get the hierarchy right and you build skills that stick. Get it wrong and you build prompt dependence that can take months to undo.
This post breaks down the four main prompt hierarchies, when to use each, and the fading principles that determine whether your client actually acquires the skill.
The Four Prompt Hierarchies
Most-to-Least (MTL)
You start with the highest level of support — often full physical guidance — and systematically reduce prompts as the client demonstrates consistent responding.
Best for: New skills where the learner needs to experience success immediately. Useful in early skill acquisition programs for clients who are still pairing and building tolerance for instruction.
Watch out for: Fading too slowly. If you stay at high-level prompts longer than necessary, MTL sets up prompt dependence quickly. Build a mastery criterion before moving to a lower prompt level — don’t fade by feel.
Least-to-Most (LTM)
You begin with the minimum level of support and increase prompts only when the client fails to respond within a set latency window. LTM gives the client an opportunity to respond independently on every trial.
Best for: Clients who have demonstrated some acquisition and need practice responding to the natural cue. LTM is also valuable for practitioners building observational skills — you learn a lot about a client’s current stimulus control by starting at the bottom.
Watch out for: Too many errors at the early prompt levels. If a client consistently fails at independent and gestural levels, LTM may be generating more error history than learning. Consider MTL instead.
Graduated Guidance (GG)
The practitioner provides physical guidance that shadows the client’s movements, removing contact the moment the client initiates the correct response. GG is continuous and responsive rather than tiered.
Best for: Motor chains and daily living skills — toothbrushing, dressing, hand washing — where you want to shape movement quality rather than step through discrete levels.
Watch out for: GG requires significant experience to execute well. Novice practitioners often either withdraw too slowly (limiting independence) or too quickly (leading to errors at critical chain links). This is one of the harder techniques to learn without direct observation and feedback.
Time Delay
Rather than varying the type of prompt, time delay varies the interval between the natural cue and a controlling prompt. In constant time delay, the interval stays fixed after a zero-delay training phase. In progressive time delay, the interval increases incrementally.
Best for: Transferring stimulus control from the prompt to the natural discriminative stimulus. Time delay works well when a client is already acquiring a skill under LTM or MTL and you want to shift control to the natural cue.
Watch out for: Clients who consistently wait for the prompt rather than responding during the delay interval. That is a sign the delay is not long enough, or that the SD is not yet functioning as a reliable cue.
The Real Skill Is in the Fade
Selecting a hierarchy is the easy part. The clinical judgment lives in how — and when — you fade.
A few principles that hold across all four systems:
- Set a criterion before you fade. Fading on instinct is how prompt dependence develops. Define the mastery criterion at each prompt level (for example, three consecutive independent responses) before moving to a lower level of support.
- Fade toward the natural cue, not just toward less support. Prompt fading isn’t only about reducing physical or gestural assistance. It means shifting control toward the discriminative stimulus. Ask whether the client is responding to the prompt or the SD — that distinction drives your decision.
- Error correction and prompt fading interact. Your error correction procedure affects which hierarchy makes sense. If you’re running a no-no-prompt procedure, LTM is often the more coherent pairing than MTL.
- Collect prompt-level data. You cannot make sound fading decisions without knowing which prompt levels are in play. If your data sheet captures only correct or incorrect rather than prompted or independent, you don’t have the information you need.
How Kipr Helps
Prompt delivery and fading are skills that develop through repetition — and in clinical practice, every trial with a real client counts. A hesitation, an over-prompt, or a fade that comes a session too late has direct consequences for that client’s learning trajectory.
Kipr lets practitioners rehearse prompting sequences with AI client personas before those decisions happen in a real session. You can run the same scenario repeatedly — practicing when to offer support, when to hold back, when the criterion is met, and how to fade — without the weight of a live therapy hour. The clinical judgment that builds slowly across years of supervised fieldwork can be developed in a safer, faster practice environment.
Join the Kipr waitlist for early access.