Push-In vs. Pull-Out Therapy for Autistic Students: What Actually Works (And Why It's More Complicated Than You Think)

Have you ever walked into a classroom to provide push-in speech therapy, only to find yourself wondering, “What am I even pushing into?”. If you're a school-based SLP, OT, or special educator supporting autistic students, you've probably been told that push-in therapy is the "gold standard." It's collaborative. It's integrative. It promotes generalization. On paper, it sounds perfect.

But push-in therapy only works when there's actually something to push into. When the classroom is chaotic, understaffed, or led by a burnt-out teacher who's barely hanging on, push-in therapy isn't integrative, it's just pull-out therapy in a louder, more dysregulating environment. That doesn't serve anyone.

In this post, I'm sharing what I learned from years of working in an approved private school with autistic students—including the good, the bad, and the messy reality of how service delivery models actually play out in real life.

What the Research Actually Says About Push-In vs. Pull-Out Therapy

Before I dive into my personal experience, let's talk about what the research tells us. Spoiler alert, both models can be effective. Studies show that push-in therapy works when services are individualized to the child's needs, there's strong collaboration between the SLP and classroom staff, and the classroom environment supports learning and engagement. Pull-out therapy works when the child needs a less stimulating environment to focus, the child needs one on one support for explicit skill building, and the child feels safe and regulated outside the classroom.

The key takeaway? There is no one-size-fits-all answer. Your service delivery model should be based on the individual child, the classroom environment, and the level of collaboration you have with the team. But here's the part the research doesn't always capture: What happens when those ideal conditions don't exist?

My Experience in Two Very Different Classrooms

Over the years, I worked in two distinctly different types of classrooms and my feelings about push-in therapy depended entirely on which type I was walking into. I had the privilege of working with two incredible teachers, Miss Krista and Miss Liz, who completely changed how I thought about collaboration and push-in therapy.

Their classrooms worked because they were organized. They had themes mapped out for the entire year, so I always knew what we were working on and could align my communication goals with their curriculum. They loved their students. I don't just mean they cared, I mean they showed up every single day ready to support, advocate, and problem-solve for their kids. They were collaborative. We planned together, communicated regularly, and supported each other's goals. When I pushed into their classroom, there was always a lesson happening, a routine in place, and a clear plan for the day. Also, they advocated for what their students needed. They even piloted a co-teaching model to see if it would better serve their students. And guess what? It did.

In this environment, push-in therapy was magic. I could weave communication goals into their lessons, model language during activities, and collaborate with the entire team. Everyone was working toward the same goals, and our students thrived because of it.

Then there were the other classrooms.

These were classrooms where teachers were chronically overwhelmed and undersupported, staff were constantly being pulled to other rooms, leaving the classroom understaffed. There was no clear structure, routine, or lesson plan. And overall, the vibe was negative, chaotic, and dysregulating. And to be clear, I don't blame the teachers. This wasn't a personal failure. It was a systemic failure. These were educators who got into the field to help kids, but were run into the ground by a system that didn't support them. They were underpaid, overworked, and given impossible caseloads with zero resources.

But because of all of these systemic problems, when I tried to push into these classrooms, there was nothing to push into. Students were sitting at their desks doing nothing. Adults were on their phones. The lights were off. No one was talking to the kids. And I was supposed to "integrate my services" into… what, exactly?

In those moments, push-in therapy wasn't effective. It was just pull-out therapy in a more dysregulating environment. And worse, I often found myself pulling students out just to give them a break from the overwhelming negativity they were surrounded by all day.

And that's not okay.

When Push-In Therapy Becomes Performative (Not Effective)

I wish more administrators understood that requiring push-in therapy doesn't automatically make it effective. If you're pushing SLPs to provide push-in services in classrooms where teachers don't have prep time, there's no collaboration happening, the classroom is chronically understaffed, and the environment is chaotic and dysregulating…then you're not improving outcomes. You're just checking a box.

And when we make do in those situations, when we try to solve the problem ourselves instead of speaking up, we miss the opportunity to advocate for systemic change. We need to be the squeaky wheel. We need to go to admin and say: "I can't provide effective push-in therapy when the teacher I'm supposed to collaborate with has no prep time and is literally using my session as her only bathroom break. How would you like me to push in under these conditions?"

Because when we stay silent, nothing changes.

How I Made Service Delivery Decisions for My Autistic Students

So how did I decide whether to push in or pull out? Honestly, I usually ended up doing a combination of both. Here's why. Pull out sessions were best for building foundational skills in a calm, regulated environment, one-on-one connection where the child could focus without distractions, explicit modeling of communication strategies the child could hear and absorb, and for understanding my students deeply (their interests, their communication styles, and their sensory needs).

Push in sessions were best for carryover and generalization of skills we'd worked on in pull-out sessions, identifying barriers that were preventing skills from transferring to the classroom, collaborating with staff to provide education and support, and observing patterns in the child's communication across different contexts.

When I used both models strategically, I could see what was working, what needed adjusting, and what barriers (sensory, environmental, or relational) were getting in the way. And most importantly, I could make decisions based on what was best for the child, and not what admin required or what looked good on paper.

The bottom line is, it's About the Child, Not the Model Here's what I want you to take away from this, push-in therapy isn't inherently better than pull-out therapy. And pull-out therapy isn't a lesser option. What matters is: individualized decision-making based on the child's needs, strong collaboration with the classroom team, a classroom environment that supports learning and regulation, and respect for the child's autonomy and sensory needs. If those conditions exist, push-in therapy can be incredibly effective. If they don't, pull-out therapy might be the best way to support your student's communication in a way that actually moves the needle forward.

And if you're being forced to push in when the conditions aren't right, it's time to speak up—not to blame the teacher, but to advocate for the systemic changes that would actually make push-in therapy effective.