I have been evaluating service dogs for public access readiness for over fifteen years. In that time I have developed a deep skepticism toward the way the industry approaches this evaluation. The standard Public Access Test is a checklist. It measures performance snapshots under controlled conditions. It does not measure the thing that actually matters, which is whether a team can function safely and reliably across the full spectrum of real environments, unexpected stimuli and human error.
The gap between passing a PAT and being genuinely ready for public access is one of the most consequential gaps in service dog training. I have seen dogs pass formal evaluations and then fail spectacularly in a grocery store. I have seen dogs get flagged during evaluations who went on to be exceptional public access partners. The instrument is imprecise and the field knows it. What I want to do here is lay out the architectural framework I use at TheraPetic® Healthcare Provider Group when I evaluate a team, why I weight certain elements heavily and why I have largely moved past treating the standard PAT as a terminal certification moment.
Why the Standard Public Access Test Falls Short
The Public Access Test as codified by Assistance Dogs International and adapted by most major credentialing bodies is a reasonable starting point. I want to be fair about that. It was designed with genuine intent, and the core behaviors it tests, loose leash walking, controlled greetings, ignoring food on the floor, load and unload from vehicles, are all real-world relevant. My critique is not with the behaviors on the list. My critique is with the evaluation architecture itself.
First, the PAT is almost always administered in a single environment on a single day. This is a fundamental measurement problem. A dog that performs beautifully in a familiar parking lot with a known evaluator on a calm Tuesday morning is not demonstrating public access readiness. It is demonstrating context-specific compliance. These are different things and conflating them is dangerous.
Second, the PAT has no meaningful measure of handler competency under stress. Most evaluations are collegial. The handler knows they are being evaluated. The evaluator is friendly. The environment is pre-scouted or familiar. None of that maps to the actual challenge of managing a service dog while simultaneously managing a medical episode, navigating a crowded hospital lobby or processing a hostile confrontation about the dog's legitimacy.
Third, the PAT is pass-fail at a moment in time. Readiness is not a binary state. It is a continuum, and the team's position on that continuum changes with fatigue, health status, environment novelty and handler emotional regulation capacity. A static snapshot evaluation cannot capture this.
Environmental Neutrality as the True Baseline
The concept I come back to more than any other when I think about public access readiness is what I call environmental neutrality. This is the dog's capacity to enter a novel environment and achieve a functional baseline of calm attention within an appropriate time window, without requiring significant handler intervention to get there.
Environmental neutrality is not the same as ignoring the environment. A dog that walks through a hardware store with glazed eyes and zero situational awareness is not neutral. That dog is potentially dissociating under stress, which is a welfare concern. A genuinely neutral dog notices novel stimuli, processes them and returns to a working orientation toward the handler. The noticing is fine. The lingering, the fixation, the inability to release from a stimulus and re-engage, that is the problem.
When I assess environmental neutrality I do not take a team to a familiar retail environment. I take them somewhere they have never been, ideally with a sensory profile that differs significantly from their training environments. I am watching entry behavior carefully. How many seconds does it take before the dog is scanning for handler cues rather than scanning the environment? What is the muscle tone at the shoulder and neck? Is the tail carriage neutral or suppressed? Is the dog offering eye contact or avoiding it?
A team that cannot achieve environmental neutrality within roughly ninety seconds of entry, in my clinical observation, is not ready for unaccompanied public access. That is not a hard cutoff written anywhere in federal guidance or ADA regulations. It is a threshold derived from fifteen years of watching teams succeed and fail in real-world conditions.
Recovery Time After Stimulus Exposure
Recovery time is the variable I weight most heavily in my evaluations. It is also the variable most consistently absent from standardized testing.
Here is the operational definition I use: recovery time is the elapsed time between a dog's peak arousal response to a stimulus and the dog's return to a pre-stimulus behavioral baseline, measured in the presence of the handler but without active handler intervention designed to accelerate recovery.
That last clause matters enormously. I want to see what the dog does on its own. Any trainer can coach a dog back to baseline if they are allowed to use their full toolkit of positioning, redirection, verbal cues and food. What I need to know is what happens when the handler is occupied, distracted or in a medical episode themselves. Can this dog regulate its own arousal response in a public environment without a trainer actively managing it?
The stimuli I use in this assessment are not the theatrical ones common to PAT protocols, the stranger approaching to pet, the shopping cart rolling by, the dropped item. Those are useful but they are too predictable. Real public access life involves a screaming child who appears suddenly from behind a display, a dog fight breaking out thirty feet away, a fire alarm, a person in a motorized wheelchair moving erratically in close proximity.
I stage these scenarios with the handler's knowledge that something unexpected will happen but without disclosing what or when. I want handler arousal elevated because real life produces handler arousal. Then I watch the dog's response curve. A peak that returns to baseline within fifteen to thirty seconds, without handler intervention beyond a neutral verbal marker, tells me this dog has genuine stimulus resilience. A peak that persists for two to three minutes, or that produces secondary behaviors like stress panting, lip licking, yawning or scanning, tells me the dog needs more proofing at this level before I can endorse unrestricted public access.
The teams I work with through officialservicedog.com Training Plus go through multiple rounds of this recovery assessment across different environments before I consider the public access readiness question answered.
Handler Focus Under Duress
Half of what I am evaluating in any public access readiness assessment is the handler, not the dog. This is where most evaluation frameworks leave enormous blind spots.
A service dog's public access performance is inseparable from handler competency. The finest trained dog on earth will develop access problems if the handler cannot read early stress signals, cannot execute a clean exit when conditions deteriorate and cannot maintain basic leash mechanics when their own nervous system is activated. I have watched otherwise excellent teams fail in public because the handler's anxiety about the environment communicated down the leash in ways the handler was not aware of.
What I assess in handlers is not confidence in the theatrical sense. Plenty of anxious handlers work beautifully with their dogs. What I assess is handler awareness, specifically the ability to observe the dog's behavioral state accurately while simultaneously navigating the environment and managing their own disability-related needs.
I use a dual-task scenario. I ask the handler to navigate a route I have given them in advance while completing a cognitively demanding secondary task, counting backward from a given number, recalling a sequence, answering questions. I watch whether leash mechanics degrade, whether the handler misses stress signals from the dog and whether the handler can execute a correction or a redirect without breaking from the secondary task.
This is not about passing or failing on clean mechanics. This is about calibrating what level of environmental complexity this team can handle right now. That information shapes my training recommendations far more precisely than any PAT score could.
What I Actually Look For in My Evaluations
When I synthesize an evaluation across the dimensions above, there are specific behavioral markers I am looking for in both members of the team.
In the dog:
- Voluntary eye contact offered at a rate appropriate to the environment's novelty level
- Loose leash maintained without handler reinforcement during low-demand periods
- Position recovery after an unexpected environmental event without handler cuing
- Suppression of investigative behavior around food, other animals and high-value environmental stimuli
- Task performance accuracy at or above ninety percent under ambient distraction
- No calming signal clusters that indicate chronic stress
In the handler:
- Accurate verbal report of dog's behavioral state when asked mid-route
- Appropriate response latency to early-stage stress signals, not waiting for a behavioral breakdown to intervene
- Clean leash transfer in tight spaces without prompting
- Ability to verbalize an exit protocol when asked hypothetically and ability to execute it when conditions require it
- Emotional regulation sufficient to avoid transmitting anxiety through leash or body language
I am not looking for perfection. I am looking for functional reliability across a realistic range of conditions. The standard the International Association of Canine Professionals and ADI both gesture toward is that the dog must not disrupt the normal course of business or pose a health or safety threat. My evaluation is designed to give me genuine confidence that a team will meet that standard on a bad day, not just a good one.
Making the Readiness Decision
The readiness decision is never made after a single evaluation session. A single session gives me a data point. I need a pattern.
My standard protocol involves a minimum of three evaluation sessions conducted across at least two distinct environmental categories, one low-complexity environment such as a quiet retail space or library and one high-complexity environment such as a transit hub, medical facility or busy food service environment. Sessions are spaced to allow for training adjustments between them. I am looking for consistency, not a peak performance on one day.
When I sign off on a team's public access readiness, I am making a professional judgment that carries real-world weight. That team is going to be in hospital waiting rooms, courthouses, classrooms and airports. Other people's comfort and safety are part of the equation. The dog's welfare is part of the equation. The handler's ability to manage their disability effectively in public is part of the equation. A checklist passed once in a parking lot cannot carry that weight.
The industry needs to move toward multi-session, multi-environment evaluation protocols with explicit handler competency components. The CCPDT framework for evaluator credentials provides a useful structural model, but the content of what we evaluate needs to expand significantly.
Public access readiness is not a graduation. It is a clinical determination about a team's capacity to function safely in a complex world. I treat it that way in every evaluation I conduct, and I have seen that rigor produce teams who succeed not just on their best days but on their hardest ones. That is the only standard worth holding.