Co-Teaching Models: 7 Proven ESL Strategies
Field Elementary School switched its ELL program from pull-out to full co-teaching and watched its students’ median growth percentile jump from “Approaching” to “Exceeds” in a single school year, a result strong enough that the district later won a state English Language Proficiency Act Excellence Award1. That kind of jump doesn’t come from swapping rooms. It comes from picking the right instructional model for the right lesson, which is the part most ESL teachers never get trained on.
What Are the Six Co-Teaching Models for ESL?
Marilyn Friend and Lynne Cook first laid out five co-teaching structures in 1995, and later work split “observing” and “assisting” into separate models, bringing the total to six2. Every push-in arrangement you’ll ever run is some version of these:
- One Teach, One Observe — one teacher leads instruction while the other collects specific data on language production, participation, or behavior.
- One Teach, One Assist — one teacher runs the full-class lesson while the co-teacher circulates and supports individual students.
- Station Teaching — the content is split into stations and small groups rotate through them, with each teacher owning one or two stations.
- Parallel Teaching — the class is split into two groups and both teachers deliver the same lesson simultaneously to a smaller audience.
- Alternative Teaching — one teacher works with the bulk of the class while the other pulls a small group for targeted pre-teaching or reteaching, still inside the room.
- Team Teaching — both teachers plan and deliver the lesson together, trading the lead role back and forth in front of the whole class.
Dove and Honigsfeld, who adapted this framework specifically for English learners, found that teams who rotated through at least three of the six models reported fewer scheduling conflicts than pairs who defaulted to one teach, one assist every single day. Variety isn’t just good pedagogy here — it’s what keeps co-teaching from collapsing into “teacher plus unpaid aide.”
Why Push-In Support Beats Pull-Out for Most ESL Classrooms
Pull-out looks efficient on a master schedule, but it tells the content teacher — and the student — that English learners don’t belong in the regular room during real instruction. Tan Huynh, who coined the shift from “push-in/pull-out” language to “sheltered co-teaching,” argues the terminology itself does damage: removing a student implies the content teacher doesn’t believe that student can handle grade-level material, and that belief tends to become self-fulfilling3.
The instructional cost is just as real. A student pulled out for 30 minutes misses whatever the class covered during that window and has to catch up later, usually without the context the rest of the room got live. Say the class spends that half-hour on a map-reading activity with partner talk built in — the pulled-out student comes back to a worksheet with no partner left to practice the language with. Push-in avoids that gap entirely because the ESL teacher is standing in the same room when the content happens, which is also the only way One Teach/One Assist or Station Teaching work at all.
None of this means pull-out is never appropriate. A brand-new arrival with zero English still needs dedicated time on survival vocabulary and basic literacy before push-in support does much good. The point is to use pull-out as a short, deliberate bridge rather than the default setting for every English learner on the caseload.
How to Set Up Station Teaching in a Mixed-Level ESL Class
Station Teaching works best when a lesson naturally breaks into distinct skills — one station for vocabulary, one for guided reading, one for independent practice. Assign each co-teacher one or two stations they’re responsible for designing, not just supervising, and rotate groups of four to six students through on a timer.
The failure mode here is giving the ESL teacher only the “support” station while the content teacher runs everything else. That’s One Teach/One Assist wearing a Station Teaching costume. Real station teaching means both teachers own content, not just one teacher owning content and the other owning management.
Parallel Teaching: Splitting the Class Without Splitting the Content
Parallel Teaching cuts your student-to-teacher ratio in half without changing what gets taught. Both teachers deliver the identical lesson to their half of the room at the same time, which is the single fastest way to get more English learners talking in a 40-minute period — half the class size means roughly double the speaking opportunities per student.
It only works if both teachers plan the lesson together beforehand, line by line. If one teacher improvises while the other follows a script, the two groups end up learning different things, and you’ve quietly created two separate classes instead of one co-taught one.
One Teach, One Assist: The Easiest Model to Start With
Most co-teaching pairs default to One Teach/One Assist because it needs the least coordination — one person leads, the other roams and helps. That’s exactly why it’s also the model most likely to turn into something unequal. When the ESL teacher spends every single lesson in the assist role, students start treating them as a helper rather than a teacher, and the content teacher stops planning with them at all.
Use this model deliberately, not by default. It’s the right call when introducing a brand-new concept that needs one clear voice, or when the ESL teacher needs a lower-prep day to catch up on the heavier models later in the week.
Alternative Teaching and Team Teaching: The Two Models Most Pairs Skip
Alternative Teaching gets overlooked because it looks like a mini pull-out, even though the small group never leaves the room. Picture a science lesson on the water cycle: the content teacher runs the main lesson while the co-teacher spends ten minutes front-loading “evaporation,” “condensation,” and “precipitation” with four newcomer students at a side table, then folds them back into the whole group once the vocabulary load drops. Same room, same lesson, just a timed vocabulary head start.
Team Teaching asks for the most trust and the most practice, which is why it’s usually the last model a new pairing attempts rather than the first. Both teachers plan every beat of the lesson together and hand the lead back and forth mid-sentence, the way two co-hosts trade a broadcast. Students stop being able to tell who’s “the real teacher,” which is the entire point — it signals that both adults in the room have equal authority over the content.
Here’s how the six models stack up when you’re choosing one for Monday’s lesson:
| Model | Best For | Prep Needed |
|---|---|---|
| One Teach, One Observe | Gathering data before planning interventions | Low |
| One Teach, One Assist | New concepts needing one clear voice | Low |
| Station Teaching | Lessons that split into distinct skills | Medium |
| Parallel Teaching | Maximizing speaking time in one period | Medium |
| Alternative Teaching | Pre-teaching vocabulary to a small group | Medium |
| Team Teaching | Established pairs modeling shared authority | High |
Co-Planning Time: The One Thing That Makes or Breaks Push-In Support
Every co-teaching model above assumes the two teachers actually talked before the lesson, and that assumption is where most push-in programs quietly fail. Without a protected planning slot, Team Teaching degrades into Team Teaching In Name Only — one person talks, the other nods — and Station Teaching collapses because nobody agreed who’s covering what content.
Even 15 fixed minutes a week beats an hour that keeps getting bumped. This is the same lesson that applies to co-teaching with local teachers in international placements: the pairs that protect a small, consistent planning window outperform the ones chasing a bigger block that never materializes.
How to Know Which Co-Teaching Model Fits Your Class
Match the model to the lesson goal instead of running the same structure every day. Introducing new grammar to the whole group calls for One Teach/One Observe or Team Teaching. Reinforcing a skill in smaller groups calls for Station Teaching. Getting every student talking in a single period calls for Parallel Teaching. Supporting a handful of SLIFE students who need pre-teaching before the main lesson calls for Alternative Teaching.
Pair this with the same differentiation thinking covered in differentiated instruction strategies for ESL — co-teaching models are really just differentiation applied to the adults in the room instead of the students. Rotate through at least three models across a two-week cycle and ask the content teacher directly which ones felt most natural; that answer usually tells you where to default on the weeks you don’t have time to plan from scratch.
Co-teaching for English learners isn’t a scheduling problem to solve once and forget. It’s a set of six tools, and the pairs who treat it that way — picking a different one depending on what Tuesday’s lesson actually needs — are the ones whose ESL teacher stops feeling like a visitor in someone else’s classroom. Pick one model you haven’t tried yet, run it for a single lesson this week, and debrief with your co-teacher for five minutes right after the bell. That’s a smaller commitment than it sounds, and it’s the fastest way to find out which of the six actually fits your classroom instead of guessing from a list.
Sources
- Co-Teaching to Support ELLs, ASCD — case study on Field Elementary School’s shift to co-teaching and its impact on ELL growth percentiles.
- 6 Models of Co-Teaching, Understood.org — overview of Friend and Cook’s original co-teaching framework.
- Rethinking Push-In Pull-Out (PIPO), Tan Huynh — argument for replacing pull-out models with sheltered co-teaching for English learners.



