The Picture Exchange Communication System (PECS) is a structured, six-phase program that teaches people with limited or no speech to communicate by handing picture cards to a communication partner. Developed in the late 1980s and grounded in applied behavior analysis, PECS is designed so the learner initiates communication rather than simply responding to prompts.1PubMed Central. Effects of picture exchange communication system on communication and behavioral anomalies in autism Each phase builds on the one before it, moving from a simple physical exchange to commenting on the world in short sentences.
Phase 1: How to Communicate
The first phase strips communication down to its most basic act: exchanging one picture for one desired item. The learner sees something they want, picks up a picture card representing that item, and hands it to a communication partner. In return, the partner immediately gives the learner the item and provides enthusiastic verbal feedback.
Two adults are involved at this stage. One serves as the communication partner, sitting across from the learner and holding the desired item. The second adult, positioned behind or beside the learner, acts as a physical prompter. If the learner does not spontaneously pick up and hand over the picture, the prompter gently guides the learner’s hand through the motion. Over successive trials, the physical guidance fades until the learner independently reaches for the card, extends it, and releases it into the partner’s open hand. The communication partner never verbally prompts the learner to “give me the picture” or “use your card.” The goal is spontaneous initiation, not obedience to a verbal cue.
Phase 2: Distance and Persistence
Once the learner reliably makes single-card exchanges at close range, Phase 2 introduces physical distance. The communication book (a binder or board holding the picture cards) is moved farther from the learner, and the communication partner gradually moves farther away as well. The learner must now get up, travel to the book, remove the correct picture, walk to the partner, and complete the exchange.
This phase teaches persistence. If the partner has their back turned or is across the room, the learner must approach, get their attention, and hand over the card. Research suggests that this second phase encourages social-communicative behavior because it requires the learner to seek out another person rather than simply complete a mechanical hand-off.2PubMed. Comparative efficacy of the picture exchange communication system (PECS) versus a speech-generating device: effects on social-communicative skills and speech development Still only one picture is on the book at a time, so the learner does not yet need to choose between images.
Phase 3: Picture Discrimination
Phase 3 is where the system becomes genuinely communicative. For the first time, multiple pictures are placed on the communication book and the learner must select the correct one. Training typically starts with two pictures: one of a highly preferred item and one of something the learner does not want. If the learner picks the preferred-item card and hands it over, they get what they asked for. If they hand over the wrong picture, the partner gives them the non-preferred item, prompting the learner to pay closer attention next time.
As the learner becomes accurate with two cards, more pictures are added. The difficulty increases further when the learner must discriminate between two or more items they both find appealing. At the end of Phase 3, a learner can flip through a book of dozens of icons, locate the one that represents what they want, remove it, and deliver it to a partner. This is the stage where the communication book begins to function like a real vocabulary.
Phase 4: Building Sentences
Up to this point, every exchange has involved a single picture. Phase 4 introduces the sentence strip, a detachable strip on the front of the communication book where the learner arranges multiple icons in sequence. The learner places an “I want” card at the left end of the strip, followed by the picture of the desired item, and then hands the entire strip to the communication partner. The partner reads the strip aloud: “I want cookie,” or “I want ball.”
Over time, additional words can be inserted. Adjectives and other descriptors are introduced so the learner might construct “I want big red ball.” The sentence strip makes the structure of a request visible and tangible, and it mirrors the left-to-right word order of spoken English. Mastery of Phase 4 is considered a significant milestone because it demonstrates that the learner can combine symbols in a structured way, a skill that carries over to more advanced communication tools later.3Perspectives on Augmentative and Alternative Communication. Strategies for Transitioning From PECS to SGD. Part I: Overview and Device Selection
Phase 5: Answering “What Do You Want?”
In the first four phases, the learner always initiates. Phase 5 introduces the first responsive exchange. The communication partner points to the “I want” icon and asks, “What do you want?” The learner then constructs the sentence strip just as they did in Phase 4. The key difference is that the learner is now responding to a direct question rather than acting entirely on internal motivation.
This phase is taught carefully so the learner does not lose the ability to initiate spontaneously. Training alternates between prompted and unprompted opportunities. A session might include moments where the partner asks the question, and other moments where the learner sees a desired item and requests it without any verbal prompt. Maintaining both spontaneous and responsive requesting ensures the learner does not become dependent on an adult’s question to communicate.
Phase 6: Commenting
The final phase moves beyond requesting and into commenting, which is communication that does not result in getting a tangible item. The learner uses picture cards to describe things, answer questions about what they see or hear, and make observations about their environment.4PubMed Central. Predictors of successful Picture Exchange Communication System training in children with communication impairments: insights from a real-world intervention in a resource-limited setting New sentence starters are introduced alongside “I want”: “I see,” “I hear,” “I feel,” “It is.”
A partner might hold up a toy airplane and ask, “What do you see?” The learner places “I see” on the sentence strip followed by the airplane icon. Or the learner might independently notice something and compose a comment without any adult prompt. Commenting is the most socially rich form of communication in the PECS protocol because it opens the door to shared attention and conversation rather than transactional exchanges.
Does PECS Lead to Spoken Words?
One of the most common concerns parents and professionals have is whether relying on pictures will discourage a child from developing speech. The research consistently points in the opposite direction. In multiple small studies, children who learned to use PECS also showed increases in spoken words. One study found that all three participating children met the learning criterion for PECS and showed accompanying increases in verbal speech.5PubMed Central. Using the picture exchange communication system (PECS) with children with autism: assessment of PECS acquisition, speech, social-communicative behavior, and problem behavior Another reported that two out of three children developed vocalization responses in the form of words and short phrases after PECS training.6Autism and Developmental Disorders. The Effect of the PECS Alternative Communication System on Vocal Speech Acquisition in Children with Autism
Researchers have also explored modifications to the standard protocol, such as having the communication partner say the word aloud as a vocal model during exchanges. In one study testing this approach, two of three participants increased their independent vocalizations after the vocal model was added.7Focus on Autism and Other Developmental Disabilities. Effects on Speech Development With Modifications to Picture Exchange Communication System The pattern across studies is not that every child gains spoken language, but that PECS training does not suppress speech and often appears to facilitate it. The picture exchange seems to pair the visual symbol with the spoken word often enough that some children begin producing the word on their own.
Effects on Challenging Behavior
Many children and adults who cannot communicate effectively express frustration in other ways: tantrums, aggression, self-injury, or simply withdrawing. Giving someone a reliable way to ask for what they want can reduce these behaviors. One case study tracked a child with autism whose aggressive behavior was maintained by wanting access to preferred items. After PECS training, the child’s vocal requests increased while aggressive behavior dropped across all three settings tested. A week after the intervention ended, the child maintained high levels of PECS use, continued using more spoken requests, and aggressive behavior remained near zero.8Behavioral Disorders. Effects of PECS on the Emergence of Vocal Mands and the Reduction of Aggressive Behavior Across Settings for a Child With Autism
The pattern holds in adults, too. A study of adults with developmental disabilities and severe communication deficits monitored off-task and tantrum behaviors during PECS training. Problem behaviors showed marked decreases, especially during the later phases of training, and remained below baseline levels during follow-up.9Remedial and Special Education. Effects of Implementing the Picture Exchange Communication System (PECS) With Adults With Developmental Disabilities and Severe Communication Deficits The reductions were most pronounced as learners progressed further through the phases and had more communicative options available to them, which makes intuitive sense: the more ways you have to tell someone what you need, the less likely you are to resort to disruptive behavior to get it.
PECS Compared With Tablets and Speech-Generating Devices
With the widespread availability of tablets and dedicated speech-generating devices (SGDs), many families wonder whether picture cards are outdated. The evidence suggests that for the specific skill of learning to request, PECS and SGDs produce comparable results. A meta-analysis comparing interventions with SGDs to other forms of augmentative communication found no differences between the two for children learning to request.10PubMed. Comparing Interventions With Speech-Generating Devices and Other Augmentative and Alternative Communication Modes: A Meta-Analysis
A head-to-head study comparing PECS with an iPad-based communication app told a similar story. All participants learned requesting skills with both systems, but all of them required more prompted trials and more sessions to reach criterion on the iPad compared with PECS.11PubMed. A Comparison of PECS and iPad to Teach Requesting to Pre-schoolers with Autistic Spectrum Disorders The tablet’s touchscreen navigation added a learning step that the picture cards did not have. Preference probes in that study were inconclusive, meaning neither system clearly won on the question of which one children preferred to use.
This does not mean one system is inherently better. PECS has a low cost of entry (laminated pictures, a binder, and Velcro) and requires no charging, Wi-Fi, or troubleshooting. SGDs offer voice output, virtually unlimited vocabulary, and growing social acceptance as tablets become ubiquitous. For many learners, the two are not in competition. PECS can serve as a starting point that teaches the fundamental concept of symbolic communication, and then the learner can transition to a device when the time is right.
When to Transition From PECS to a Device
Clinicians generally consider a learner ready to transition to an SGD after mastering Phase 4, when they are independently combining an “I want” symbol with vocabulary icons on a sentence strip and using a large, categorized vocabulary.12Perspectives on Augmentative and Alternative Communication. Strategies for Transitioning From PECS to SGD. Part I: Overview and Device Selection At that point, the learner has demonstrated the ability to combine symbols into sentences, which is exactly what they will need to do on a device’s grid-based display.
The transition process is not simply swapping one tool for the other. It begins with an assessment of the learner’s current PECS skills so that the device can be set up to mirror their existing vocabulary and sentence structures. Teaching strategies from the PECS protocol, like prompting and fading, carry over. The minimum criterion for considering the transition complete is that the learner uses the SGD appropriately and independently across environments, and their communication with the device matches or exceeds what they achieved with PECS.13Perspectives on Augmentative and Alternative Communication. Strategies for Transitioning From PECS to SGD. Part 2: Maintaining Communication Competency If the learner’s communication regresses after introducing the device, the team may need to step back, simplify the device layout, or return to PECS temporarily while retraining specific skills.
Using PECS With Peers
Standard PECS training involves an adult communication partner, but communication in daily life often happens between same-age peers. Researchers have adapted PECS protocols to include typically developing peers as communication partners, with promising results. In a peer-mediated intervention with minimally verbal preschoolers, all four children with autism improved in peer-directed communication. Two children expanded beyond requesting and began using pictures to comment and share, and all four showed improved social engagement. Their peers also increased their communication with the children with autism.14PubMed Central. Picture Exchange Communication System and Pals: A Peer-Mediated Augmentative and Alternative Communication Intervention for Minimally Verbal Preschoolers With Autism
Another study taught children with autism to direct their picture exchanges toward peers rather than adults. Independent requests directed at peers increased, and the peers responded with brief social interactions following most exchanges.15PubMed Central. Teaching children with autism to engage in peer-directed mands using a picture exchange communication system The finding matters because a child who only communicates with therapists and parents misses out on the peer interactions that drive social development in school and playground settings. Training peers to be communication partners is a relatively simple extension of the standard protocol and can make the skill more functional in the environments where the child actually spends their day.
Beyond Autism
PECS was originally designed for children with autism, and the majority of published research focuses on that population. But the system has been used with people who have a range of communication challenges, including intellectual disabilities, Down syndrome, cerebral palsy, and acquired communication disorders. A quantitative synthesis of single-participant studies found that individuals without autism diagnoses actually showed slightly higher levels of requesting (manding) when taught PECS, though the difference was not statistically significant.16Remedial and Special Education. Quantitative Synthesis and Component Analysis of Single-Participant Studies on the Picture Exchange Communication System
The same synthesis found that PECS taught in a single setting produced somewhat higher requesting levels than PECS taught across multiple settings, though again the difference was not statistically significant. This pattern might reflect the added complexity of generalizing a new skill across different rooms, different partners, and different routines all at once. Practitioners working outside of autism contexts can use PECS with reasonable confidence that the phase structure transfers, though they should keep in mind that most of the evidence base still comes from the autism literature.
How Training Quality Shapes Outcomes
The effectiveness of any structured communication program depends heavily on whether the person implementing it follows the protocol correctly. A coaching study examined what happened when therapists received targeted instruction on delivering PECS. After coaching, therapists increased their use of the AAC strategies they had been trained on, and their clients increased their use of communication in those trained contexts. However, both therapists and clients showed limited generalization to untrained contexts.17PubMed. Impacts of a PECS instructional coaching intervention on practitioners and children with autism
That last detail is worth sitting with. If a therapist learns to support PECS exchanges during snack time but never practices doing so during free play, neither they nor the child will naturally carry the skill into free play without additional effort. Generalization across settings has to be planned and practiced, which is why Phase 2’s emphasis on varying distance and partners is not just a training exercise but a design principle. The more environments and partners a learner practices with during each phase, the more portable the skill becomes. Families and school teams who embed PECS practice into mealtimes, car rides, and playground routines tend to see broader use than those who confine practice to a therapy room.
What PECS Does Not Do
For all its strengths, PECS has boundaries. It teaches functional communication, primarily requesting, with commenting entering only in the final phase. It does not teach conversation in the back-and-forth, turn-taking sense. It does not teach literacy, though some learners begin to recognize printed words placed alongside the icons. And it does not replace speech therapy for learners who have the potential to develop spoken language; it works best as a complement, giving the learner an immediate way to communicate while speech and language skills develop in parallel.
The system is also inherently limited by its physical format. A learner who uses a large communication book must carry it everywhere, keep it organized, and have access to new icons whenever their vocabulary needs to expand. In practice, a book that works perfectly in a classroom may be left in a backpack at recess or forgotten at home during a weekend outing. These logistical challenges are part of what drives the eventual transition to an electronic device for many learners, since a tablet can hold thousands of symbols without weighing any more.
Finally, not every learner progresses through all six phases. Some plateau at Phase 3 or Phase 4, and that outcome is still meaningful. A child who can independently select a picture of their preferred snack from a book of 30 icons and hand it to a caregiver has gained functional communication even if they never reach commenting. The phase structure provides a roadmap, but every learner travels it at their own pace and may stop at a point that meets their daily communication needs.