An ABA mastery criterion is the rule, written before teaching begins, that tells the team when a child has learned a skill well enough to change the teaching. The rule usually pairs an accuracy level with a number of observations, such as 80% correct across three consecutive sessions. A good criterion does more than close out a goal. It protects the child from being moved on too early, and it protects the family from a goal list that looks finished on paper but does not hold up at home.

This guide explains how criteria work, shows examples you will see on a treatment plan, and covers what research says about higher and lower standards. It also covers when a criterion should be changed and which questions to ask your child’s team.

TL;DR

• A mastery criterion is a measurable rule, set before teaching, that says when a skill counts as acquired. It usually combines an accuracy level with a number of consecutive sessions.
• Common examples run from 80% to 100% accuracy across two or three consecutive sessions. Research suggests higher criteria can support better maintenance for some learners.
• Meeting a criterion does not prove the skill will last or show up with new people and places. Maintenance and generalization need their own checks.
• A criterion should be revised when data show skill loss, weak generalization, or heavy prompting. It should never be loosened just to declare success.
• Parents can ask three questions: what is the rule, how is it measured, and what happens after the goal is met.

Table of Contents

• What a mastery criterion actually is
• Examples of ABA mastery criteria
• How teams choose a criterion for one child
• What research says about higher and lower criteria
• How teams know the criterion has been met
• Mastered does not mean maintained or generalized
• When a criterion should change
• A note from our clinical perspective
• Questions to bring to your child’s team
• How we can help
• FAQ
• Sources

What a mastery criterion actually is

In ABA, a mastery criterion is a predetermined accuracy or performance standard that signals a skill has been acquired. It is set before instruction so the decision is objective. The team is not guessing, and no one is relying on a good day. Researchers describe this rule as having at least two dimensions: the performance level, such as percent correct, and the number of observations or consecutive sessions at that level.

The criterion also tells the team what happens next. Once it is met, the team can fade prompts, introduce new targets, or begin checking whether the skill holds over time. Without a stated rule, goals can drag on for months, or they can be marked complete after a single strong session.

Here is a plain example. Suppose a goal reads: “When shown a picture of a common object and asked ‘What is it?’, Maya will name it.” The criterion might be: 90% correct across two consecutive sessions, with different therapists, without prompts. Each part of that sentence matters:

• The level (90%) sets how accurate the child must be.
• The count (two consecutive sessions) makes sure it was not luck.
• The conditions (different therapists, no prompts) begin to build in generalization.

Our own guide to how skills count as mastered walks through this from the family side.

Examples of ABA mastery criteria

There is no single universal number. Fuller and Fienup (2018) note that research and practice commonly use criteria from 80% to 100% accuracy, often across two or three consecutive observations. Studies of skill maintenance report examples of 80%, 90%, and 100% correct responding, with skills taught until the criterion was met and then probed weekly for four weeks.

The table below shows how those building blocks look on a plan. These are illustrative formats, not prescriptions for any particular child.

That last row is worth a moment. Not every criterion is a percentage. For skills like greeting a peer or washing hands, the standard is often “independent across people and settings.” For a parent, that format is usually the most meaningful, because it describes what you will see at the grocery store, not just at the therapy table.

A worked example

Say a child is learning to request a break by handing over a card. Sessions have 10 opportunities each.

1. Session one: 6 of 10 independent requests (60%).
2. Session two: 8 of 10 (80%).
3. Session three: 8 of 10 (80%).
4. Session four: 9 of 10 (90%).

Under a rule of “80% across three consecutive sessions,” the goal is met after session four, because sessions two, three, and four were all at or above 80%. Under “90% across two consecutive sessions,” it is not met yet, and teaching continues. Same child, same data, different decision. That is why the rule needs to be visible and agreed on before the first session.

How teams choose a criterion for one child

A criterion should fit the learner and the skill, not a template. Carlson and colleagues describe clinical recommendations of 80% to 100% correct for two or three consecutive sessions, while stressing that criteria should be adapted to individual needs. Recent work on individualized measurement also points out that clinicians can select a performance criterion based on the level of true mastery they want to infer and how confident they need to be, taking the number of trials into account.

That last idea is easy to miss. A score like 8 out of 10 is an inference. Ramos (2025) makes the distinction between the observable performance criterion and true mastery: it is evidence about an underlying probability of responding correctly, not a guarantee that the child will be correct every time. This is why a single session at 80% rarely settles anything, and why more observations make a decision sturdier.

A few questions help a team set the number:

• How important is the skill? A skill tied to safety, such as stopping at a curb, deserves a stricter standard than a preference for sorting colors.
• How many opportunities does each session offer? Ten trials give a steadier picture than three.
• Has the child shown trouble keeping skills in the past? If so, a higher level or more sessions may be wise.
• What will the skill look like in real life? If the target is conversation, the criterion should reflect more than flashcard accuracy.

If your child is working on conversation, our parent guide to conversation skills practice shows how that kind of goal moves from structured practice into daily talk.

What research says about higher and lower criteria

Most of the evidence here comes from small studies, so it is best read as guidance and not as a rule. Still, the pattern is consistent enough to take seriously.

Fuller and Fienup compared 50%, 80%, and 90% criteria with three children with autism. The 90% criterion produced higher maintenance responding three to four weeks after teaching ended. The authors call for more research, and their view is that the criterion is not merely an administrative cutoff: the level used to end teaching has predictive value for later maintenance.

Another line of work found something similar. Anderson and colleagues report that the most commonly surveyed practitioner choice was 80% accuracy across three consecutive sessions, yet their experiments found this may be insufficient for maintenance in some cases. And in a study of four children aged five to nine learning word recognition, 100% mastery produced higher maintenance accuracy than 80% or 90%. The authors read this as support for higher criteria, including 100% across three sessions, when the clinical priority is durable responding, and they still stress adapting to the learner and skill.

Researchers and clinicians also tend to differ in style. A descriptive analysis of the literature found that applied researchers more often used higher accuracy across fewer observations, while clinicians more often used lower accuracy across more observations. Neither approach is wrong in itself. The point for families is that two reasonable teams can write different rules, so ask why yours chose the one it did.

How teams know the criterion has been met

The rule only works if the data behind it are sound. One practical issue is how much of a session gets recorded. A classic comparison found that recording every instructional trial is more sensitive to changes in performance, while first-trial-only data can lead to premature mastery decisions. The authors recommend continuous recording when feasible because it better represents what the child is doing.

For parents, that translates into a simple question: “Do you record every trial, or a sample?” Neither answer is automatically a problem, since sessions are busy, but the team should be able to explain their method and why it is enough for the decision at hand.

Two other habits make mastery calls more trustworthy:

• Look at the trend, not one point. A child who goes 60%, 80%, 80%, 90% tells a different story than one who goes 100%, 40%, 100%, 60%.
• Check that the people measuring are doing it the same way. Consistent delivery of teaching steps matters, which is why we pay attention to treatment fidelity.

For a plain-language tour of what gets recorded and why, see our guide on how behavior analysts collect data, and our piece on reliable progress tracking for separating real change from session-to-session noise.

Mastered does not mean maintained or generalized

This is the part families most need to hear. Meeting a criterion in the teaching setup does not show that the skill will persist after instruction ends or occur with new people, materials, cues, or places. Those need separate probes, and reviews show they are often left out.

A scoping review of ABA for children and youth with autism found that 41% of records included some follow-up measure, 40% included a mastery or criterion measure, and 31% included a generalization measure. The authors also caution that reaching a mastery criterion, especially a low one, may not demonstrate maintenance.

What does good practice look like after mastery? Maintenance can involve stopping instruction, teaching less often, or probing at intervals to see whether more teaching is needed. A simple pattern on a plan might read:

1. Criterion met in the clinic or home with the primary therapist.
2. A probe with a second adult and different materials.
3. A probe in another room or community setting.
4. A maintenance check a week later, and again after several weeks, with no teaching in between.

The professional standard supports this thinking. The BACB Ethics Code expects interventions to be based on assessment results and designed to produce outcomes likely to maintain under naturalistic conditions. A criterion that includes later maintenance and real-life contexts is a direct way to honor that. Our clinic-to-home generalization checklist gives caregivers and clinicians a way to plan the handoff.

When a criterion should change

Criteria are not carved in stone, but they should never be moved casually. A team should reconsider the rule when data show any of these:

• The skill drops during maintenance checks.
• It does not carry over to new people or settings.
• Progress has stalled over a reasonable stretch of time.
• The child is leaning on heavy prompting to reach the number.
• The child’s needs or environment have changed, such as a new school or a new sibling at home.
• The standard is too low for how important the skill is, which can be a safety concern.

The Ethics Code points the same way: behavior analysts are expected to collect and use data appropriately and continually monitor and evaluate their interventions, selecting assessments on scientific evidence and client needs. In practice, a revision should be written down, tied to ongoing data and assessment results, and made with the family’s input. The change should serve the child’s outcome in daily life. It should not be a way to declare a goal finished.

A useful test: if a criterion is being lowered, can the team explain, in one sentence, how that helps the child and not the chart?

A note from our clinical perspective

We have spent more than two decades watching what happens after a goal is checked off. The children who thrive are not the ones who hit a number fastest. They are the ones whose teams treated the number as the start of a conversation.

When we write a criterion, we want it to be clear enough that a parent could read it and know exactly what success looks like. We also want it to point past the therapy table: to a second person, a second room, and a check a few weeks later. Our approach is built around individualized plans and parent coaching because skills hold best when the people in a child’s everyday life know what to look for and how to support it.

If a plan you are given says only “80%” with nothing about who, where, or how long, ask us or any provider to fill in the rest. That is a reasonable question, and a good team will welcome it. Mastery is a hopeful milestone, and it deserves a plan that carries it into real life.

Jennifer

Questions to bring to your child’s team

You do not need clinical training to ask good questions. These keep the conversation concrete:

• What is the exact criterion for this goal, including level, number of sessions, and conditions?
• How is it measured, and are all trials recorded?
• Who needs to see the skill before we call it mastered?
• What happens after mastery: when are maintenance and generalization checked?
• Under what circumstances would you change the criterion, and how would I hear about it?

Parents sometimes worry that asking these questions will seem like second-guessing. It is the opposite. Teams work better when families understand the rules, and parent involvement is one of the strongest supports for skills carrying over to the home. Our guide on parent coaching sessions shows what that partnership can look like week to week.

How we can help

At Building Block Resolutions, every treatment plan is individualized, and mastery criteria are part of that design: written clearly, shared with caregivers, and linked to what we want to see outside of sessions. We offer ABA autism therapy, parent coaching so that goals are supported at home, and remote ABA therapy for families who prefer to work from home.

If you would like to talk through your child’s current goals, or learn how we set and review criteria, reach out through our main site to book an appointment. We can explain our process and what to expect in the first weeks. For a broader overview, start with our clear guide to ABA therapy for parents.

FAQ

What is the most common ABA mastery criterion?

One of the most common is 80% accuracy across three consecutive sessions. A survey of practitioners found it was the most commonly used, though the same researchers found it may not be enough for maintenance in some cases. Many plans use 90% or 100% for skills that must be especially durable.

Does meeting a mastery criterion mean my child has mastered the skill for good?

Not automatically. Meeting the rule shows the skill is reliable under the teaching conditions. Research reviews note that maintenance and generalization require separate probes. Ask your team how and when they will check the skill with new people, places, and materials.

Can a mastery criterion be changed?

Yes, when the data and the child’s needs call for it. Reasons include skill loss at maintenance checks, weak generalization, or heavy prompting. The change should be documented and based on ongoing data, as the BACB Ethics Code expects of continual monitoring, and it should not be made simply to declare success.

Is a higher criterion always better?

Not always. Some studies, including one with 100% versus 80% and 90% criteria, found better maintenance with higher standards. The same authors stress fitting the criterion to the learner and the skill. A very high bar for an unimportant skill can cost teaching time, so the team should weigh both.

How many trials do clinicians record to decide mastery?

Ideally every instructional trial. A comparison of recording methods found continuous data are more sensitive to performance changes, while first-trial-only data can lead to premature decisions. Your team can explain which method they use and why it suits the goal.

Sources

• A Preliminary Analysis of Mastery Criterion Level: Effects on Response Maintenance
• Mastery Criteria and Maintenance: a Descriptive Analysis of Applied Research Procedures
• Mastery criteria and the maintenance of skills in children with autism
• The effects of different mastery criteria on the skill maintenance of children with developmental disabilities
• A Comparison of Methods for Collecting Data on Performance During Discrete Trial Teaching
• Applied Behavior Analysis in Children and Youth with Autism Spectrum Disorder: A Scoping Review
• MIEBL: Measurement of Individualized, Evidence-Based Learning
• Ethics Code for Behavior Analysts (BACB)

Recommended

• Goal Mastery Criteria in ABA: How Skills Count as Mastered
• ABA Progress Tracking: Spot Reliable Data, Not Session Noise
• IEP Goals With ABA: How to Write Ones That Work