Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the reactive dog research overview.
Short answer. A dog under threshold can learn; a dog over threshold cannot. The veterinary behaviour evidence (AVSAB, 2021; Overall, 2013) is consistent that reward-based work only changes the underlying emotional response when the dog is in a physiological state where learning is possible — and that state is defined by distance, intensity, and exposure duration to triggers. Staying under threshold is therefore not a stylistic choice in fear and reactivity work. It is the mechanism.
Karen Overall, in Manual of Clinical Behavioral Medicine for Dogs and Cats (2013), describes the threshold as the point at which a dog moves from reactive but recoverable into a sympathetic-arousal state where executive function is offline. Below threshold, the dog can take food, orient to the handler, perform known cues, and form new associations between a trigger and a positive outcome. Above threshold, the dog is in fight-or-flight: heart rate elevated, pupils dilated, food refused, learning suspended. Each over-threshold reactive episode practices the reactive response and strengthens the neural circuitry behind it (Overall, 2013).
Grisha Stewart's Behavior Adjustment Training 2.0 (2016) describes the same construct from a training-protocol angle: under threshold is where functional reinforcers (distance, sniffing, choice) work; over threshold is where they don't. The two frames converge on the same operational rule — keep exposures inside the window where the dog's nervous system is still regulated.
Threshold is not a fixed property of a dog; it varies by trigger, context, day, and cumulative stress load. The practical question is not "what is my dog's threshold?" but "what is my dog's threshold for this trigger, today, after the morning we just had?"
The body-language markers that precede an over-threshold reaction are well documented in the canine behaviour literature (Yin, 2009; Overall, 2013; Mills et al., 2013). The earliest reliable signs:
These markers appear before the bark, lunge, or growl. The window between the first marker and the over-threshold reaction is often only 1–3 seconds in a reactive dog — but it is enough time, with rehearsal, to add distance.
The single most useful practical reframe in reactive-dog walks is this: walk length is the wrong unit. Distance under threshold is the right one. A 60-minute walk in a busy park where the dog is over threshold three times is worse than a 20-minute walk in a quiet alley where the dog stays calm throughout. The longer walk feels like more "work" and more enrichment; in fact it has trained the reactive response three times.
The McConnell / Donaldson tradition (McConnell, 2002; Donaldson, 1996) emphasises this same point — that the goal of any walk during the behaviour-modification phase is not to expose the dog to triggers in order to "socialise it" or "get it used to" them. Flooding above threshold does not produce habituation in fear-driven cases; it produces sensitisation (AVSAB, 2021). The goal is to manage exposure so that every trigger encounter happens under threshold, where counter-conditioning can do its work.
Practical planning rules drawn from this framing:
1. Choose routes with sightlines. A long straight street with visibility 100m ahead lets you see triggers early and add distance. A blind corner does not. 2. Match time-of-day to your dog's trigger profile. Most reactive dogs have predictable bad windows — school pickup, commuter dog-walking peaks, evening rush. Walking outside those windows is not avoidance; it is management. 3. Treat distance as cheap. Crossing the road, doubling back, ducking behind a parked car — these are not failures. They are the protocol working. 4. Carry food the dog will actually take in mild stress. Dry kibble is not enough for most reactive dogs; high-value protein (chicken, cheese, freeze-dried liver) is the working currency.
Every reactive-dog handler will misjudge threshold. The honest expectation in the literature is not perfection but a low rate of over-threshold incidents and a quick, calm recovery from the ones that happen (AVSAB, 2021). The recovery script that consistently appears across the IAABC and Fear Free practitioner literature:
1. Add distance immediately. Turn, move, do not wait to see if the dog recovers in place. 2. Do not correct, jerk the leash, or apologise to the other handler. Each of those adds stimulus to a dog who is already over threshold. 3. Once calm, drop several pieces of food. The marker is the dog re-orienting to you, not the trigger disappearing. 4. End the walk earlier than planned. A reactive episode raises baseline cortisol for hours to days; stacking more triggers on top extends recovery (Mills et al., 2013).
The veterinary behaviour evidence is unambiguous on what to avoid in threshold work: aversive corrections (leash pops, prong, e-collar) at or near threshold (Ziv, 2017; AVSAB, 2021); flooding (forced exposure above threshold in the hope of habituation); and the "let her work through it" school, which mistakes suppression for learning. In fear-driven reactivity, suppression of the visible bark and lunge often leaves the underlying fear intact or worsened — and produces dogs who go from "no warning" directly to bite (Overall, 2013).
Most reactive dogs improve substantially over 6–24 months of consistent under-threshold work (AVSAB, 2021). The visible signal is rarely the bark going away first. It is the threshold distance shrinking — 50m to 30m to 15m for the same trigger — and recovery time shortening from minutes to seconds. These are the two markers worth tracking, because they move long before the dog "looks normal" to a casual observer.
---
Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full reactive dog research overview for the complete framework.