StarHealing
Article

How many ABA hours per week does an autistic child actually need?

What the intensity literature actually compares, what our own milestone records show for four families, and how to budget hours a real household can keep.

Parents are usually told "25 to 40 hours a week" in a single breath, and then left to work out what that means for a household where both parents work. The honest answer is that intensity is a range, not a target, and the ceiling is set by your family rather than by the literature.

Where the 40-hour number came from

The high-intensity recommendation traces back to a single 1987 study with a small sample and a very specific setup: clinic-delivered sessions, trained supervisors, near-daily hours. Later replications produced mixed effect sizes, and the wider reading of that literature is that the first hours matter much more than the later ones. The gap between 15 and 30 hours is smaller than the gap between 0 and 15.

That asymmetry is what you should plan against. Buying your fifteenth hour is a different decision from buying your twenty-eighth.

What we can show you from our own families

We publish our own records rather than a target. In the Stage 1 pilot and its longitudinal follow-up, four families enrolled and three completed the first milestone; two of those children continued and were assessed at four milestones between June 2025 and April 2026. Over that period one child's DSM-5-domain total fell from 15 to 8, moving from Level 2 to Level 1, and one child began attending a mainstream kindergarten.

Both write-ups are OSF preprints — Stage 1 and follow-up — not peer-reviewed papers, with no control group and mostly caregiver-reported measures. They are four families, not a trial. What they do show is the shape of our week: we run structured work in the range of a dozen to fifteen hours, and the rest of the child's day is family life, not therapy.

How to spend fifteen hours well

  • Table-top teaching is the smallest share, not the largest. A few hours, not most of the week.
  • Play with siblings or peers generalises in a way a desk cannot.
  • Parent-implemented blocks, even twenty minutes a day, carry more than an extra clinic hour.
  • Nap time and car rides are not therapy time. Rest is a variable in the outcome.

When fifteen hours is not possible

Take what your household can hold for six months rather than what looks correct for one week. Dropping to zero after three exhausted weeks is a worse outcome than a steady eight hours. Consistency across weeks is the variable we can see in our own follow-up data: the two children who kept going are the two children whose later milestones improved.

FAQ

Is 10 hours a week of ABA enough?

It is enough to produce measurable change, and the difference between 10 and 15 hours is smaller than the difference between 0 and 10. What you protect across weeks matters more than the weekly total.

Does AI or a screen replace clinic hours?

No. In our programme the AI layer does progress tracking, session planning and report writing, which frees clinician hours for direct teaching. It does not substitute a device for a person.

How fast should I expect change?

We organise progress in milestones rather than weekly targets. The first milestone in our published records took roughly a three-month stretch of regular sessions. Look for a specific skill you can name by the end of the first milestone; if nothing you can name has shifted by then, the plan needs redesigning, not more hours.

Why do some scores go up while the child gets better?

Because a measure can only score what it can observe. A child who was not yet speaking could not be rated on language-related items at all. When speech appears, those items become scorable and the total can rise even though the child has gained a capacity. We explain this in full on our method page.