When does therapy become daycare? The 40-hour question in ABA

Put two schedules side by side. Same time clock, same 40 hours. One is an adult's work week. The other is a three-year-old's therapy schedule. Forty hours is also about how long a typically developing child of that age might spend in daycare.

I work in ABA, and I believe in it. So this is a question I am asking from inside the field: at what point does a comprehensive schedule stop being therapy and start being supervised care? In this post I trace where the 40-hour number came from, what the newest and largest studies say about hours, what the federal government's own regulator just wrote, and the one question I want you to ask before you accept a schedule.

If your child has been recommended 30 or 40 hours and you are not sure it fits your family, a 45-minute family care mapping call can help you lay out the whole picture: school, therapies, rest, and family time. Book a 45-minute family care mapping call

I am a BCBA, but I am not your BCBA. This is general education, not treatment, and not a treatment recommendation for your child or your state. Build the real plan with the people who actually treat your child.

Where 40 hours came from

Every conversation you have had about a comprehensive, 40-hour week traces back to one paper. In 1987, Ivar Lovaas at UCLA reported on 19 children who received 40 hours a week of one-to-one behavioral treatment for two years or more. Nine of the 19, or 47%, reached what the study called normal intellectual and educational functioning by age seven or eight, compared with 2% of a comparison group that did not get that intensity. A 1993 follow-up found eight of those nine had held that status.

Here is the part that gets skipped in every casual retelling. This was one high-intensity group compared against a lower-intensity group. It was not a dose-response study. Nobody tested whether 20 hours would have gotten those same children to the same place. Nobody tested 30. Modern reviews also point out real problems: no random assignment, techniques that included aversives nobody in this field would defend today, and a narrow, contested definition of "normal functioning." The study never answered the question of how much. It just kept getting cited until insurance would pay for it.

What the guidelines actually say

The Council of Autism Service Providers (CASP) 2024 practice guidelines are the field's closest thing to a consensus standard. They define two tiers, not one number: focused treatment at 10 to 25 hours a week, targeting a limited number of goals, and comprehensive treatment at 26 to 40 hours a week, addressing multiple developmental domains. They say hours should be individualized to the child's actual need, not defaulted to the top of either range.

The Behavior Analyst Certification Board, the body that certifies me, does not require any specific number of hours. On paper, this was never supposed to be one size fits all. So why do we behave like it is?

What the research says about hours

A lot of content about ABA gets this wrong in both directions. A 2010 meta-analysis of 22 studies found a dose-dependent effect: more hours tracked with better outcomes when the data was pooled. That effect did not hold consistently within individual studies.

Then came the largest dataset anyone has looked at on this question: Sandbank and colleagues in JAMA Pediatrics, 144 studies and 9,038 children. It found that outcomes did not improve as intensity increased. In the lead author's words, there was not rigorous evidence that increasing the amount of intervention produces better outcomes.

In 2026, real-world data from the Catalight Research Institute on 725 children added nuance. More hours predicted more goals attained, but more hours were negatively associated with progress on adaptive behavior and communication, the skills most parents would call the whole point. Reductions in dangerous behavior were unrelated to hours. The authors concluded that practice guidelines should temper their language around dosage. If someone tells you the research proves your child needs more hours, they have not read the research.

What the federal regulator wrote

In August 2026, the Centers for Medicare and Medicaid Services (CMS) released a state Medicaid and CHIP toolkit on ABA. It reports the national average at 17.33 hours per beneficiary per week in 2025, up 22% since 2021. Medicaid and CHIP ABA spending rose 421% from 2021 to 2025, from $1.94 billion to $10.1 billion, while the number of children receiving ABA grew 189% and autism diagnoses grew 67%. Spending is growing far faster than the population it is supposed to track.

And then the sharpest line in the whole episode, which does not come from a critic. CMS writes that 40 hours of ABA per week is not a best practice. It says services at higher intensity levels should not be authorized by default, and that high-intensity weekly schedules "leave virtually no time for non-treatment activities like schooling and family time and can easily take on the characteristics of respite care and child care." CMS also notes there is no consensus on a set number of recommended hours anywhere in the field.

Trying to fit ABA, speech, OT, and school into one week? We can map it together. Book a 45-minute call

Close to home: North Carolina

North Carolina, one of the states I am licensed in, passed House Bill 696 in 2026. Treatment plans over 16 hours a week now need state or managed care approval on a much shorter leash, with a review every three months. The scale behind it: North Carolina's Medicaid ABA spending went from $1.9 million in 2020 to $505 million in fiscal year 2025. We opened the floodgates for access, providers came running, and now access is being choked. It is a familiar movie.

Continuity needs a workforce

Intensive ABA was built as a two-person job from the start: a graduate-trained analyst to assess and supervise, and a trained paraprofessional to sit with the child most of the hours. Insurance mandates made the hour payable, and the RBT credential made that paraprofessional billable and auditable. The workforce exploded at the technician tier, because that is where the paid hours are.

Median RBT turnover runs about 65% a year, and some large providers report 89% to over 100%, replacing more than their entire frontline staff annually. Hospital staff turnover is around 21%. The 2025 BCBA exam pass rate fell to 51%, the lowest on record, while BCBA job postings far outnumber practicing BCBAs. A high-hour schedule assumes the same faces and a clinician who knows your child well enough to individualize the plan. The current system makes that continuity structurally hard, not because the people in it do not care.

My take: therapy, school, respite, and daycare are different jobs

These are my opinions, not facts, and you are welcome to disagree. Therapy, by definition, is a prescribed dose aimed at a specific goal, measured, and meant to taper down. School is a day program. Daycare is a trusted option when school is not the right fit. Respite exists so exhausted parents can sleep. An ABA clinic billed as therapy is being asked to be all of those at once, and it cannot be.

Meanwhile, families are carrying the rest. When institutions closed, the funding per person that followed people into home and community services was a fraction of the institutional cost, and parents became the unpaid workforce: driving to every appointment, translating between providers who never talk to each other, and getting up at night. What many families actually need is a break, and that is not built into the current model of support.

There are also financial incentives worth naming. Fee-for-service pays by the hour, so more billable hours means more revenue, whether or not those hours are clinically justified for that specific child. Early intervention clients are easier to get approved for high hours than older children who may need targeted help just as much. I have seen companies rely on the hours for their bottom line. Autistic self-advocates, a Harvard public health researcher, and CMS are all describing the same pattern from different rooms. That is not noise.

The fair version of both sides

Position one: ABA's evidence, taken as a whole, supports high-intensity comprehensive treatment for children with significant needs. CASP already calls for individualization. The failure is implementation and enforcement, so payers should audit for defaulting to the maximum instead of capping hours for everyone.

Position two: fee-for-service billing structurally rewards maximum billable hours, and the newest and largest studies increasingly undercut the idea that more hours reliably produces better outcomes. Under this reading, defaulting to comprehensive is what the incentive structure predictably produces, whether or not any one clinician intends it. Both positions point to real evidence. And we still do not know how many children leave ABA because they met their goals versus because of cost, access, or burnout. One Kaiser Permanente study found that of children who started ABA, 66% were still in services at 12 months and 46% at 24 months. That is retention, not success, and that gap should bother our field more than it does.

The question to ask

The 40-hour number is not settled science. It is one 1987 study that hardened into an industry default, running into a workforce that may not be able to deliver it safely even when it is clinically justified. Our children are missing school, missing other therapies, and sometimes losing progress while they wait out a billing cycle with us.

Here is your concrete next step. The next time a schedule is set at comprehensive, ask what it is specifically targeting, and why that number for your child. Do not accept the default just because it is the default. And ask the bigger question with me: are we the medical model, or are we specialized child care? If it is the second, families deserve to have that named and funded honestly.

Tell me in the comments: how many hours a week was your child recommended, and did anyone explain why that number?

Sources

• Sandbank et al. (2024), Determining associations between intervention amount and outcomes for young autistic children, JAMA Pediatrics
• CASP ABA Practice Guidelines (2024)
• Lovaas (1987), Behavioral treatment and normal educational and intellectual functioning in young autistic children
• CMS, Medicaid and CHIP resources on applied behavior analysis
• Autistic Self Advocacy Network
• BACB

Draft sources list. The E019 description's sources are still pending the EDIT-PLAN-v2 section 6 check; Virues-Ortega (2010), the Catalight 2026 study, the CMS August 2026 toolkit, NC HB 696, and the Kaiser Permanente retention study once confirmed.

Ashleigh Dominguez, MS, BCBA, is a Board Certified Behavior Analyst licensed in North Carolina, Georgia, and Iowa. This article is general education based on the episode. It is not clinical advice, not treatment, not legal advice, and not a treatment recommendation for your child or your state. Ashleigh is not your BCBA and has not met your family. Please build your real plan with the providers who know your child.