Do horse-based therapies help autistic children and teens?

What 8 sources say about riding lessons, hippotherapy and ground-based horse work — and what they can’t yet tell you.

Research synthesisUpdated 2026-10-038 sources includedProtocol v4.5 (Amdt. v4.6)
Therapeutic riding Promising but contested
Hippotherapy Not enough research yet
Ground-based horse work Not enough research yet

A care note before you read. There are three different kinds of horse programs. They are run by different people and aim at different things, and what we found for one kind does not tell you about the others. Horses are large animals, and programs often check whether a child can safely take part. Please talk with your child’s care team and the program about your own child.

Podcast·Two-voice deep dive

Listen to the discussion

0:00
Chapters & transcript

This is an AI-generated discussion of this article, transcribed automatically. The written summary below is the reviewed source — if the two differ, go by the summary.

0:00

What is this discussion about?

Host A What if I told you that one of the most promising interventions being studied for autistic kids right now isn't happening in some sterile clinic with fluorescent lights and clipboards?

Host B Yeah, no clipboards here.

Host A Right. It's actually out in a dusty barn involving a living, breathing, totally unpredictable 1,000-pound animal.

Host B Which is wild, right? I mean, it completely upends how we usually think about clinical data.

Host A Exactly. Normally, behavioral research is locked down tight. Like every variable is controlled. You sit at a desk, perform a task, and they measure the output.

Host B Yeah. So introducing a massive animal and an outdoor farm setting, no less, into a protocol, it's just the absolute definition of injecting wild variables into an experiment.

Host A Definitely.

Host B It forces the researchers to get incredibly creative with how they measure success.

Host A Well, welcome to this deep dive, if you're listening. Today, we are exploring equine-assisted therapies.

Host B Specifically, horse-based programs for autistic children and teens.

Host A Yep. We're looking at a pretty fascinating stack of research today to figure out, you know, what actually happens when you bring these kids into the equestrian environment.

Host B We've got a lot of ground to cover.

Host A We really do.

Host B We're pulling from broad overviews that map the whole landscape of equine therapy. So we've got a 2017 systematic mapping review by McDaniel Peters and a 2023 meta-analysis by Xiao.

Host A And we also have some really granular data.

Host B Right, from a major 2015 randomized control trial by Gabriels, plus a 2020 single-case experimental design study. Oh, and some audit documents. And we're even going to fact-check a Wikipedia page about all this today.

Host A Which is going to be interesting.

Host B That combination, you know, it gives us a really unique vantage point.

Host A Because we get the big picture and the details?

Host B Exactly. We can see the 30,000-foot view of the field, but we can also zoom all the way in on how, say, a specific 10-week riding program impacts a single group of kids.

Host A But before we get into the weeds, I do want to state plainly that the mission of this deep dive is strictly to provide a summary of research findings, not advice.

Host B Right. Very important distinction.

Host A Absolutely. Nothing in this discussion is a recommendation for any individual child. We are just here to look at the data, period.

Host B Just the science.

2:20

What are the three kinds of horse therapy?

Host A Okay, let's unpack this phrase, horse therapy, because, I mean, people picture wildly different things when they hear that, right?

Host B Oh, totally.

Host A Like, I might picture a kid learning to trot around a ring while someone else pictures, I don't know, just petting a pony.

Host B Right.

Host A And the literature is actually incredibly strict about separating these activities.

Host B So how do they break it down?

Host A Well, the sources divide the landscape into three very distinct categories, and they're delivered by totally different professionals targeting entirely different outcomes.

Host B Okay, what's the first one?

Host A First, you have therapeutic horseback riding, or THR. This is essentially adapted riding lessons. The primary activity is learning actual equestrian skills.

Host B Okay.

Host A It's led by a certified riding instructor, and they're usually flanked by volunteers who walk alongside the horse just to ensure safety.

Host B So the kid is like actively holding the reins, learning to steer, learning to stop.

Host A Yes.

Host B The child is in control of the animal.

Host A Got it. And the second one?

Host B The second category is hippotherapy. And this operates on a completely different medical model. How so? Here you have a licensed physical, occupational, or speech therapist. And they are using the horse itself as a treatment tool.

Host A So they aren't learning to ride?

Host B No, absolutely not. The child is mounted on the horse, sure, but they aren't learning to ride at all. The therapist is actually manipulating the horse's movement to achieve a specific physiological or speech goal.

Host A Oh, wow. Okay. And the third category strips the riding out entirely, right?

Host B Exactly. Okay. That is ground-based work. In the literature, you'll sometimes see it called occupational therapy in an equine environment or equine-assisted learning.

Host A But no riding.

Host B Zero riding. The child stays on the ground the entire time. They're grooming the horse, leading it by a halter, maybe cleaning stalls, and just learning general animal care.

Host A Okay. So it feels like the difference between taking a driving lesson, which would be the therapeutic riding, and doing physical therapy while sitting in the passenger seat of a moving car, which is hippotherapy, and then taking an auto mechanic class in the garage, which would be the ground-based work.

Host B That's, yeah, that is a great analogy. It perfectly separates the fundamental goals of the three. Though, I mean, to be fair to the biomechanics of hippotherapy, sitting in a passenger seat doesn't exactly engage your core the way riding a horse does.

Host A Fair point, yeah. Car seats are a bit more comfortable.

Host B Right. A horse's pelvis actually perfectly mimics the three-dimensional movement of the human walking gait.

Host A Wait, really?

Host B Yeah. So when a physical therapist puts a child on a horse, that multidimensional rhythmic motion is actively working the child's trunk, their nervous system, sensory processing in a way a car seat just never could.

Host A Ah, so the movement of the animal is literally the medicine in that context.

Host B Exactly.

Host A That makes a lot of sense. And across all three of these totally distinct approaches, the programs claim they can target things like irritability, social communication, language development and motor skills.

Host B That's what the claims are. Yeah.

5:28

How much research is there, and who did it?

Host A So since we have three different interventions happening in the same barn, how did the evidence base for all this actually develop?

Host B Well, the research basically followed a classic progression, right, from broad observation to targeted testing.

Host A Like a proof of concept at first.

Host B Exactly. The 2017 mapping review we're looking at, it basically gathered 33 studies just to see what was even out there. And it was mostly small, single group studies or case reports.

Host A But it grew from there.

Host B It did. Over time, the volume grew to include pooled data, like the 2023 meta-analysis we mentioned, which combined 25 studies and 639 participants.

Host A That's a decent size.

Host B Yeah. But the absolute anchor of this evidence base, like the highest quality data we have, is one major randomized controlled trial testing therapeutic riding from 2015, plus one small single-case experimental design study testing the ground-based work from 2020.

Host A Okay, wait, looking at the authors on these papers, the 2017 mapping review and the 2020 ground-based study are both authored by someone named Peters.

Host B Yep.

Host A I mean, when I see the same name mapping the whole field and then publishing the specific trials, it kind of makes me wonder, are independent teams actually replicating these findings?

Host B No, they are not. And the sources actually highlight that as a limitation. Independent teams have not widely replicated the original work in the documents we retrieved.

Host A So it's a pretty small circle.

Host B Very small. The primary randomized control trial for therapeutic riding is essentially owned by one specific research team. And the ground-based study comes from the same author who mapped the field. So, yeah, we are relying heavily on isolated pockets of researchers.

7:13

Which children were studied, and for how long?

Host A Okay, but knowing who is running the trials is really only half the battle. I'm honestly fascinated by the logistics of who actually gets to participate.

Host B It's a huge factor.

Host A Because, I mean, you can't just take any child, regardless of their behavioral profile or physical needs, and just drop them onto a massive animal, right? There has to be a screening process.

Host B Oh, absolutely. And that screening process aggressively shapes the data. The sources note that these programs have very strict eligibility requirements.

Host A Like what?

Host B Things like weight limits, behavioral safety, meaning no severe aggression or self-injury that could spook the animal, and medical contraindications, like certain spinal issues.

Host A Okay, so they're filtering people out from the start.

Host B Exactly. So the autistic children and adolescents in these studies, who generally range from 3 to 16 years old, they represent a highly screened-in slice of the broad autistic population. Not everyone.

Host A Right. We're looking at a subset of kids who are already physically and behaviorally capable of safely navigating a farm environment.

Host B It's a crucial caveat to keep in mind.

Host A So what about the timeline and the size of these studies?

Host B Well, if you look at the major 2015 RCT by Gabriels, they enrolled 116 children at baseline, specifically ages 6 to 16, who passed those screenings.

Host A And how long were they riding?

Host B They ran the therapeutic riding intervention for exactly 10 weeks.

Host A What about the other study?

Host B The smaller ground-based study also ran for exactly 10 weeks, but it only looked at six youths.

Host A Just six. Wow. Okay. But 10 weeks is a pretty standard intervention window. What exactly are they measuring, though, when a kid is just riding a horse for an hour a week?

Host B In the primary RCT, the main outcome they cared about was irritability.

Host A Irritability?

Host B Yeah. And they measured this using standard behavioral checklists filled out by the caregivers. They also tracked social communication, hyperactivity, and spoken language, which they measured through actual language samples.

Host A Wait, how does simply riding a horse translate to a kid being less irritable at home or talking more? Is it like the core strength required or maybe the rhythmic motion is just soothing their nervous system?

Host B Or maybe just the confidence of controlling a giant animal.

Host A Yeah, exactly.

Host B The literature hypothesizes a mix of all those mechanisms. Like, the rhythm of the horse's gait might regulate the nervous system, which could bring a hyper-aroused child down to a baseline where they can finally focus.

Host A That makes sense.

Host B And the core strength they develop might translate to less physical discomfort in their daily lives. Plus, the motivation of bonding with an animal creates this highly engaging environment to practice communication.

Host A So they actually want to talk.

Host B Right. A child who might not want to talk to a clinician in some fluorescent room might be incredibly motivated to say go or whoa to a horse.

Host A Which brings up the most frustrating question in all short-term research, I think. What happens at week 11?

Host B Ah, yes.

Host A Did anyone follow up afterwards? Did they check on these kids a month or a year later to see if that nervous system regulation actually held up?

Host B Unfortunately, no. None of the retrieved studies followed up afterward. The data basically just falls off a cliff at week 10. We have absolutely no idea if the effects last once the child stops going to the barn.

Host A Man, that is frustrating. Okay, so we have a 10-week window, highly screened kids aged 6 to 16, measuring irritability, hyperactivity, and communication with zero long-term tracking.

Host B That's the setup.

10:49

Did riding lessons help irritability and talking?

Host A Here's where it gets really interesting. Let's dig into the actual numbers. What did the 2015 therapeutic riding trial actually find?

Host B Well, they found distinct, statistically significant positive results, and these started to emerge around week five.

Host A Okay, what kind of results?

Host B Caregivers reported reduced irritability with an effect size of 0.50 and reduced hyperactivity with an effect size of 0.53.

Host A So for our listener trying to visualize that, an effect size around 0.5 is a moderate difference. It's not, you know, a complete behavioral transformation, but it's highly noticeable in day-to-day life.

Host B Exactly.

Host A It might mean a child transitions from school to home without a major meltdown, or maybe they can sit through a family dinner when they couldn't before.

Host B That is exactly the real-world translation. And they also saw improvements in social communication with an effect size of 0.63. Wow. Plus an increase in total words spoken and new words spoken. Both of those showed an effect size of 0.54.

Host A An effect size of 0.54 in spoken language is massive for a minimally speaking child. I mean, that is the difference between pointing at the fridge in frustration and verbally asking for water.

Host B It really is.

Host A That is functional independence being built on the back of a horse.

Host B It's a profound finding for that specific subgroup.

12:11

What about ground-based work and hippotherapy?

Host B But, you know, we do have to look at the whole picture.

Host A Right, the other studies.

Host B Yeah, the small ground-based study, that one had mixed results. So four out of the six participating youths saw a decrease in irritability, while two didn't.

Host A Okay. What about the big meta-analysis?

Host B The 2023 pooled meta-analysis showed broad improvements in social cognition and communication. But importantly, it explicitly reported null results for things like social awareness, mannerisms, and stereotypy.

Host A So the horse doesn't cure repetitive behaviors or mannerisms.

Host B Right.

Host A Which honestly makes sense. But what about hippotherapy? You know, the physical therapy using the 3D pelvic movement of the horse. We haven't heard any results for that.

Host B And we won't. We hit a complete data void in our sources for hippotherapy.

Host A Wait, really? Nothing?

Host B Well, studies on it definitely exist in the world, but the specific database queries used to build our source stack only captured their titles via a system called CrossRef.

Host A Oh, I see.

Host B So without the full texts, there is simply not enough research here to report any findings, positive or negative, for hippotherapy.

13:19

Why does one source say there is no evidence?

Host A Okay, I want to pivot to a really bizarre contradiction in the stack then. Because we just talked through a published randomized control trial showing moderate, life-altering effects for irritability and language.

Host B Right.

Host A But the Wikipedia page included in our sources explicitly states there is no evidence that therapeutic horseback riding is effective in treating children with autism. How does that happen?

Host B Well, tertiary sources like Wikipedia often lag behind primary literature. Or they rely on outdated, highly rigid medical reviews that just dismiss anything that isn't a pharmaceutical-grade double-blind trial. It really highlights why you have to read the actual studies.

13:59

Is it safe? Were falls or injuries reported?

Host B But to your point on contradictions, there's another glaring anomaly in the data regarding safety.

Host A Yes, I was going to mention that. Across all these studies involving thousand-pound animals, dirt, outdoor farm equipment, zero sources reported any adverse effects, dropouts, falls, or harms. Zero. I mean, I find that incredibly hard to believe.

Host B As you should. In clinical research, a finding of not reported absolutely does not mean the rate of harm is zero.

Host A Right. It just means they didn't write it down.

Host B Exactly. It just means the researchers didn't prioritize capturing or publishing the safety data. And when you have a complex physical intervention and the safety data is entirely blank, well, it forces clinicians to look very critically at the foundation of the evidence.

Host A So let's do that. Let's poke holes in that foundation.

14:49

Why is the evidence rated low?

Host A Officially, the sources classify the evidence for therapeutic riding as low certainty.

Host B Yes. Promising, but contested.

Host A And the main reason for that low certainty rating seems painfully obvious when you look at the study designed for irritability.

Host B You're talking about the unblinded caregivers?

Host A Exactly. I mean, if the primary measure of success is whether the child is less irritable, and the person filling out that rating scale is the parent who drives the kid to the barn every Tuesday.

Host B Right.

Host A They watch them smile on the horse. They desperately want the therapy to work. That parent is massively biased. They know their kid is in the riding group, not the control group. Doesn't that skew the data? It's like placebo by proxy.

Host B This raises an important question.

Host A And you're completely right. Unblinded parent ratings fundamentally weaken the certainty of behavioral trials. You just cannot separate the genuine effect of the horse from the parent's optimism.

Host B Yeah, it's impossible. However, the trial design actually has a brilliant counterweight to that bias, which is why the study isn't just thrown out entirely.

Host A Oh, wait. The language samples.

Host B Yes.

Host A Yes.

Host B The researchers recorded audio clips of the children speaking. Those audio files were then sent to independent raters who simply counted the words and categorized them.

Host A Oh, I see where this is going.

Host B Yeah. Those raters were totally blinded. They had absolutely no idea if the child in the audio clip had been riding horses for 10 weeks or had never even seen a barn.

Host A That is so smart.

Host B Because the blinded, objective measure of spoken words showed a positive effect alongside the unblinded parent ratings. It anchors the study in objective reality.

Host A That is such a clever design choice. But the sources do flag other strict limitations, don't they?

Host B Oh, definitely. For one, the 2023 meta-analysis took a fairly dangerous statistical risk. The authors pooled different delivery forms together, mixing therapeutic riding, hippotherapy, and groundwork into one massive dataset.

Host A Muddying the waters.

Host B Exactly. When you mix a driving lesson, physical therapy, and auto mechanics into the same spreadsheet, it becomes nearly impossible to know which specific activity is driving the positive results.

Host A Yeah, that's a mess.

Host B Furthermore, in the 2017 mapping review that looked at those 33 early studies, only three of those studies actually confirmed the autism diagnoses of the participants using standard clinical tools.

Host A Wait, what did the rest of them do?

Host B They just relied on parent report or previous chart notes.

Host A That is wild. Okay, so we have unblinded parents, a muddy meta-analysis, missing diagnostic rigor in the early days, and zero independent replication of the big RCT.

Host B It's a lot of limitations.

Host A We also have a mountain of things the science has just completely missed. If we frame these strictly as gaps in the science, not as, you know, caution, advice, what are the most glaring open questions here?

Host B We already mentioned the lack of long-term data. We have no idea what happens at week 11 or a year later. But demographic representation is another massive gap. Like who? Adults. Adults are completely unrepresented in these findings. Every data point focuses on children and adolescents.

Host A What about things like sensory processing or motor skills? You always hear those pitched as the main benefits of being on a horse.

Host B The sources state that the data collected on motor skills, sensory processing, and even family stress levels was simply insufficient to draw any conclusions.

Host A Oh, really?

Host B The evidence just isn't there yet to support those specific claims. And most critically, like we talked about, the actual rates of falls, injuries, or horse-related incidents specifically for autistic children in these programs never measured or captured. That remains a huge open question.

Host A Okay, so if I'm a professional staring at this mountain of caveats, you know, promising language bursts, unblinded parents, missing safety data, how does a clinician actually evaluate a body of evidence like this without getting totally overwhelmed?

Host B Well, clinicians reviewing this literature rely on a very systematic reasoning process. First, they categorically refuse to lump the interventions together.

Host A So they don't let the meta-analyses muddy the waters for them.

Host B Exactly. They treat a win for therapeutic riding as completely isolated from hippotherapy or ground-based work.

Host A Right.

Host B Second, they look at the concept of the estimand. They look at how researchers control the environment. Because creating a placebo for a horse is incredibly difficult. You can't give a child a sugar pill that feels like a thousand-pound animal. Obviously not. But in the 2015 Gabriels trial, the researchers designed a phenomenal active control group. The kids in the control arm didn't just sit at home. They traveled to the facility and engaged in a whole curriculum of barn activities without the horse.

Host A Oh, so they controlled for the farm itself. Because otherwise you wouldn't know if the child was improving because of the horse or just because they got to leave the house, breathe fresh air, smell the hay, and get attention from an instructor.

Host B Precisely. By making the control group do horse-free barn activities, they isolated the horse as the active ingredient.

Host A That is brilliant.

Host B And that active control group is what builds immense confidence for clinicians looking at this data. They weigh that active control and the blinded word counts heavily in favor of the intervention's promise.

Host A But they still have concerns.

Host B Oh, definitely. Their confidence is weakened by the unblinded parent ratings. And to properly assess feasibility, they would absolutely need to see safety and injury rates strictly measured.

Host A So what does this all mean?

Host B Okay.

20:22

What are the key takeaways?

Host B If I'm a listener trying to synthesize everything we've unpacked today, here are the absolute essential takeaways based on the sources. Go for it. First, 10 weeks of therapeutic horseback riding resulted in improved irritability, hyperactivity, and social communication in one large, well-designed trial.

Host A Yep.

Host B Second, the three forms of horse programs are distinct. Riding, hippotherapy, and ground-based work operate on different models. The evidence for one does not apply to the others, and hippotherapy and groundwork simply do not have enough research yet.

Host A Third, a big limitation. The evidence for riding is officially graded as low certainty. It's heavily impacted by the fact that parents knew which group their child was in when rating their behavior, and the trial hasn't been independently replicated.

Host B And finally, the remaining questions. We have zero published data on the actual safety or injury rates, nor do we know if the benefits last after the 10-week program ends.

Host A Beautifully summarized. Now, decisions about any individual child belong in a conversation with that child's pediatrician or care team, and this deep dive cannot make that call. Research on this topic continues, and Spectrum Connect will update its coverage as new evidence appears.

Host B Always changing.

Host A Spectrum Connect translates autism research into plain language for parents and checks findings for accuracy against misinformation. If you are a credentialed expert, MD, PhD, PsyD, or DO, and would like to help validate our findings, email research@spectrumconnect.us.

Host B You know, looking at how hard those researchers had to work to isolate the horse from the farm environment, it leaves me with a final thought to mull over.

Host A Oh, what's that?

Host B Well, if an active control group doing horse-free barn activities is what proved the animal was the active ingredient, it makes you wonder, what other elements of a farm environment, just the quiet, the smell of the hay, or the wide open outdoor space, might be deeply therapeutic on their own, even without the 1,000-pound animal in the room?

Key Takeaways

Riding lessons look promising, from one large fair test. It had 116 children aged 6 to 16. Half had 10 weeks of riding lessons and half did barn activities with no horse. The riding group had less irritability and hyperactivity, said more words, and did better on social communication, with gains starting by week 5.

It is only one test, by one research team. Parents filled in the irritability forms and knew which group their child was in, which can make results look better. The word counts were scored by people who did not know the groups, which helps. We rate our certainty as low.

Not enough research yet for hippotherapy or ground-based horse work. For hippotherapy we found a review and several small studies, but could only see their titles. For ground-based work there is one small study of 6 youth, mostly based on what parents reported.

Falls and injuries are not well checked. None of the studies we could read reported falls or injuries. That does not mean there were none, and we did not find injury numbers for autistic children.

If you are thinking about one, ask the program. Which of the three kinds is it? Who leads it and what training do they have? How does it check whether your child can safely take part? What goals will it track, and what does it cost?

What this means for you

“Horse therapy” covers three different things. Therapeutic riding is riding lessons led by an instructor trained for disability, often with helpers walking beside the horse. Hippotherapy is a physical, occupational or speech therapist using the way a horse moves as part of the child’s therapy; the child is on the horse but is not learning to ride. Ground-based work is grooming, leading and caring for a horse without riding, and some of it is led by an occupational therapist.

Only riding lessons have a large fair test behind them, and it found less irritability, more talking and better social communication after 10 weeks. We call the result “contested” for two reasons. It rests on one test, with parents rating irritability while knowing which group their child was in. And a general reference page says there is no evidence that therapeutic riding works for autistic children, which conflicts with that test; we could not trace the studies that page cites. A 2023 review gathered 25 studies of all three kinds mixed together and reported gains in some areas and none in several others, so it cannot tell us which kind caused the gains. This page covers horses only. Dog programs are covered in a separate article, and reviews that mix many animals together are not used to judge horse programs.

Riding lessons look promising in one large fair test. For hippotherapy and for work with horses on the ground, there is not enough research yet to say.

Who was studied. Children and teens aged 6 to 16 in the riding test, who had 10 weeks of riding lessons; the 2023 review covered mainly ages 3 to 16 across all kinds of horse programs; the ground-based study had 6 youth, with a no-treatment baseline and then 10 weeks of sessions. Horse work screens children in or out, so the results apply only to children who passed a program’s screening; the one screening protocol we saw was piloted in the small ground-based study, and the criteria are not recorded in our notes. We did not find results for adults or for effects that last after the program ends. We did not find cost or insurance information; horse programs can cost money, take travel time, and may not be close to home. Where the studies took place is not captured in our record. We read the riding test’s abstract, not the full paper, and the 2023 review through a tool that summarizes web pages, so some exact numbers are missing. We did not search the full medical databases. Whether these results apply to any particular child or family is not established.

Where the studies landed

Promising but contested (riding lessons only)

2 Points toward it helping

2 Mixed or inconclusive

4 Context only: not read, or not a study

Tap the cover to open it. The hippotherapy cover stands for five small studies we could see only by title. A grant record and a 2018 veterinary-journal article surfaced as titles only and are not shown. Reviews that mix many animals together are not used to judge horse programs.
One studyReview of many studies

Tap any tile to read it. The hippotherapy tile stands for five small studies we could see only by title. A grant record and a 2018 veterinary-journal article surfaced as titles only and are not shown. Reviews that mix many animals together are not used to judge horse programs.

Points toward it helping2
One fair test (Gabriels 2015) of 10 weeks of riding lessons in 116 children found less irritability and hyperactivity, more words and better social communication than barn activities with no horse. A small study of ground-based work led by an occupational therapist (Peters 2020, 6 youth, no comparison group) found that parents reported more goals met and more social interest, and four of the six were less irritable.
Mixed or inconclusive2
A 2023 review (Xiao) of 25 studies and 639 people reported gains in social skills, communication, irritability and hyperactivity but none in several other areas, and it mixed riding, hippotherapy and ground work, so we cannot say which kind caused the gains. A 2017 mapping review (McDaniel Peters & Wood) of 33 studies describes what research exists, calls the field early, and gives no effect size.
Context only: not read, or not a study4
Two reviews (Atun Einy & Lotan 2017 on hippotherapy, Cleary 2024 on mental health and well-being) and five hippotherapy studies were seen only as titles or publication records, so we cannot tell you what they found. A general reference page says there is no evidence that therapeutic riding works for autistic children; it conflicts with the one fair test, and we could not trace the studies it cites. None of these are counted toward a result.

Spectrum Connect reviews published research on interventions parents are exploring for their autistic children — so you can see where the evidence actually stands. No agenda, no selling, no cherry-picking. Just the studies, our method, and what it means for you.

See the research behind this For anyone who wants to check our work · 8 sources
01

Where we looked

This was a web-based scoping search, not a full multi-database search: 2 web searches plus 10 page-read attempts, of which 8 returned content. A publisher page for one study could not be read in full (access refused), and a link for another study redirected to a different site and was not followed, so its record was read from the Crossref metadata index instead. No database was searched directly, and PubMed and PMC pages were not used, so there are no per-database counts. The search of record for a full run covers PubMed/MEDLINE, Embase, APA PsycINFO, CINAHL, ERIC, SPORTDiscus, Cochrane CENTRAL, Epistemonikos, trial registers (ClinicalTrials.gov, ICTRP), HABRI Central and Crossref update records. Dogs and other animals are left out at screening, not in the search string. Here is the PubMed search of record.

("Equine-Assisted Therapy"[MeSH] OR "Animal Assisted Therapy"[MeSH] OR hippotherap*[tiab] OR "therapeutic riding"[tiab] OR "therapeutic horseback riding"[tiab] OR equine*[tiab] OR horse*[tiab]) AND ("Autism Spectrum Disorder"[MeSH] OR autis*[tiab] OR asperger*[tiab]) Run on PubMed →

Counts for every database, and the dates searched, are pending. This search is not reproducible from the record we kept, and no protocol was registered in advance.

02

What we did with what we found

30records surfaced: 18 web results and 12 Crossref records for “hippotherapy autism”
about −5duplicate copies of the same web works
—judged by title, not formally screened
Set asideprovider, clinic and advocacy pages, mirror copies, 9 Crossref titles (no count kept)
10listed in the record, 8 shown here (1 fair test + 1 single-case study + 4 reviews + 1 general reference page + a group of 5 hippotherapy studies seen by title)

This is a scoping count, not a formal screening flow. No full texts were read: one publisher page was read through a tool that summarizes web pages, three works from their abstracts, two reviews from their publication record only, and one general reference page. The other two of the 10 listed items (a grant record and a 2018 journal article) surfaced as titles only and are not shown.

03

What the strongest evidence says

Riding lessons: irritability and hyperactivity

Less irritability (effect size 0.50) and less hyperactivity (0.53) in one fair test of 116 children, compared with barn activities with no horse. Parents rated irritability and knew which group their child was in.

How sure
Low
Riding lessons: talking and social communication

More words spoken (0.54), more new words (0.54), better social communication (0.63) and social cognition (0.41) in the same test. Word counts were scored by raters who did not know the groups. One test by one team.

How sure
Low
All three kinds mixed together (2023 review)

25 studies and 639 people. Gains in social cognition, communication, irritability and hyperactivity; none in social awareness, mannerisms, motivation, lethargy, stereotypy or inappropriate speech; too little data on movement and senses. Not credited to any one kind.

How sure
Not rated
Hippotherapy

A review and several small studies exist, but we saw only their titles, so we cannot say what they found.

How sure
Not enough research
Ground-based work with an occupational therapist

One small study of 6 youth with no comparison group: parents said more goals were met and children were more social, and four of the six were less irritable. Mostly parent report.

How sure
Not enough research
Harms: falls, injuries and horse-related incidents

None of the studies we could read reported them, and none gave written safety rules, but not reported is not the same as none. We found no injury numbers for autistic children.

How sure
Not tracked
Ray Kawai · Protocol v4.5 BCAT · Open record
Spectrum Connect is not a medical provider, and nothing here is medical advice. This page shows where the research stands and how we got there. It is not a recommendation, and it is not a substitute for your child’s doctor or care team. What you do with it is yours to decide, together with them.

What would you like to do?

Want to talk to a parent who’s been there?

We match you with families whose kid is like yours. Free, takes two minutes.

Join Free
Heard of something we haven’t covered?Suggest an article

If another parent, a clinic or a video mentioned a therapy you can’t find here, tell us. We’ll read the research and review it next.

Suggest an article
Think we got something wrong?Report a factual error

We publish the whole record so it can be checked, and that only counts if we act on what you find. If a number looks wrong, a study is missing or has been retracted, or we’ve read a finding in a way the evidence doesn’t support, tell us.

You don’t need a research background to file one. “This doesn’t match what our doctor told us” is a useful report. Every one reaches a person: we reply within seven days, and within thirty we have either corrected the page or told you when we will. Substantive reports send the affected steps back through the protocol and need fresh sign-off before anything here changes.

Report a factual error
Are you a clinician or researcher?Review this article

If you have read this and the evidence holds up, you can sign off on it. We send you a personal review link where you confirm your credentials, and our team checks every review before anything is published. Once approved, your name and a link to your profile appear on this article.

Ask to review this