Differential Reinforcement — DRA, DRI, DRO, and DRL Explained for Pet Behavior Modification

What Differential Reinforcement Means

Differential reinforcement is one of the cornerstones of applied behavior analysis, and it is one of the most useful tools a force-free trainer can deploy for problem behaviors. The phrase sounds technical, but the core idea is simple: rather than trying to suppress a behavior you do not want, you systematically reinforce something else — something that competes with or replaces the unwanted behavior. The unwanted behavior, no longer reinforced, fades; the new behavior, increasingly reinforced, takes its place. Applied behavior analysts, including Susan Friedman in her widely taught Living and Learning with Animals curriculum, place differential reinforcement at the center of humane behavior change.

There are four major flavors of differential reinforcement, each named for what you reinforce relative to the problem behavior: DRA (alternative), DRI (incompatible), DRO (other), and DRL (lower rate). Choosing among them is a clinical decision that depends on what the animal is doing, what triggers it, and what consequence has been maintaining it. The framework also sits inside Susan Friedman’s LIMA hierarchy as the third tier — the positive-reinforcement-of-alternatives tier — which most credentialed mainstream behavior consultants consider the workhorse of ethical behavior modification.

DRA — Differential Reinforcement of Alternative Behavior

DRA reinforces a chosen alternative behavior in place of the problem behavior. The alternative does not have to be physically incompatible with the problem behavior; it just has to be something the animal can do instead, and something the handler decides to pay for. Classic example: a dog who jumps on visitors. DRA replaces jumping with sit. Sit is paid for richly; jumping is no longer paid for (no eye contact, no touch, no greeting). The dog learns that sit is the more reinforced option and offers it preferentially. Our greeting-without-jumping protocol uses DRA as its mechanical core and walks through the consistency requirements that keep the contingency clean.

DRA is the most general of the four procedures and the easiest for handlers to implement once they understand the principle. The critical detail is that the alternative behavior must be reinforced at a high enough rate, with a high enough value, to outcompete whatever was reinforcing the problem behavior. If jumping was getting greeting and sit is getting only verbal praise, the contingency is unbalanced and DRA will not produce reliable change. Match the reinforcer to what the animal actually wants.

DRI — Differential Reinforcement of Incompatible Behavior

DRI is a tighter version of DRA. The reinforced alternative is physically incompatible with the problem behavior — they cannot both happen at the same time. A dog cannot be lying on a mat and jumping on a visitor simultaneously. A dog cannot be holding a stationary nose-touch on a target and barking through the window. A cat cannot be eating from a puzzle feeder and biting the resident dog’s tail. DRI uses physics to enforce mutual exclusivity. When DRI works, it works dramatically because the structural conflict between the two behaviors does some of the trainer’s work.

The classic DRI application in dog training is the down-on-mat protocol for door barking. The dog is reinforced for lying on a designated mat near the door. Lying on the mat is physically incompatible with running to the door and barking. With enough reinforcement of the mat behavior, the dog defaults to the mat when the doorbell rings. DRI also features heavily in resource-guarding protocols, where the dog is reinforced for behaviors that physically conflict with the guarding posture. Our resource guarding protocol integrates DRI with careful antecedent management and counter-conditioning, showing how DRI nests inside larger behavior-change plans.

DRO — Differential Reinforcement of Other Behavior

DRO reinforces the absence of the problem behavior during a defined time interval. The animal is paid for not doing the problem behavior — regardless of what they are doing instead, as long as it is not the targeted problem. The procedural definition is interval-based: every thirty seconds, every two minutes, whatever the interval is set at, the animal earns a reinforcer if they have not performed the problem behavior during that interval. The interval lengthens as the behavior improves.

DRO can be a useful tool but it has subtle traps. The first trap is that DRO does not teach the animal what to do, only what not to do, and animals can fill the void with new and creative problem behaviors. The second trap is that DRO is easy to mis-time. If the animal performs the problem behavior at second 29 of a thirty-second interval and you reinforce at second 30, you have just reinforced something close to the problem behavior. Best practice is to use DRO as a supplement to DRA or DRI, not as a standalone, and to set intervals conservatively so the animal can succeed. DRO has its place in research and in some clinical applications, but most pet-side practitioners reach for DRA or DRI first.

DRL — Differential Reinforcement of Lower Rate

DRL reinforces reductions in problem-behavior frequency, without requiring zero. The handler accepts that the behavior will continue at some baseline rate and pays the animal for performing it less often. DRL is useful when a behavior cannot ethically or practically be eliminated entirely. A dog who barks ten times during a one-hour training session can be reinforced for sessions in which they bark eight times, then six, then four. The criterion is rate, not absence.

DRL fits well when complete extinction of a behavior is not realistic or desirable. Some barking is normal dog behavior. Some scratching is normal cat behavior. A dog who jumps occasionally on family members but never on strangers may not need zero jumping — they may need a reduced frequency. DRL works on the rate dimension and gives the handler a tool for shaping toward an acceptable level without setting the bar at perfection.

How to Choose Among DRA, DRI, DRO, and DRL

The choice depends on the behavior, the trigger, and the handler’s resources. DRA is the default starting point for most pet-side problem behaviors because it teaches a replacement skill cleanly. DRI is preferred when a physically incompatible alternative exists and can be reinforced reliably — mat work for door barking, station work for counter surfing. DRO is a niche tool for situations where defining a clean alternative is hard, often used alongside another procedure. DRL is the procedure of choice when the goal is rate reduction, not elimination, and when complete absence is not realistic.

The clinical reasoning behind the choice is part of the ABC framework — antecedent, behavior, consequence — which we cover in our ABC functional analysis article. A clean ABC analysis tells you what is reinforcing the problem behavior, and that information drives the choice of differential reinforcement procedure. Without that analysis, picking among the four procedures is guesswork.

Differential Reinforcement Inside the LIMA Hierarchy

Susan Friedman’s LIMA hierarchy — Least Intrusive, Minimally Aversive — places differential reinforcement at tier three, after wellness and medical workup (tier one) and antecedent arrangement (tier two). The order matters. Before you reach for any behavior-change procedure, you rule out pain, thyroid dysfunction, neurological causes, and sensory loss as drivers of the behavior. Then you change the environment so the behavior is less triggered. Only then do you reach for differential reinforcement as the active intervention. Our LIMA hierarchy article walks through the full tier structure and explains why most credentialed behavior professionals consider differential reinforcement the workhorse of ethical behavior change.

Pairing Differential Reinforcement with Counter-Conditioning

Differential reinforcement is operant — it shapes voluntary behavior. Counter-conditioning is classical — it changes the underlying emotional response. For most behavior problems, especially those with a fear or arousal component, you need both. The dog who reacts to other dogs needs an alternative operant behavior (look-at-handler, eye-engagement, mat-stationing) AND a changed classical-conditioning history with the trigger (other-dog now predicts good things, not threat). The integrated approach pairs the operant differential-reinforcement plan with classical counter-conditioning for the emotional substrate.

Pairing Differential Reinforcement with the Premack Principle

The reinforcer in a differential reinforcement procedure does not have to be food. Premack reinforcers — access to high-probability activities the animal already wants — can carry the contingency just as effectively, often more effectively in environments where food has lost salience. A dog being trained on DRI mat work near the front door can be reinforced with the door opening when the mat behavior is held. A dog being trained on DRA loose-leash walking can be reinforced with sniffing access. Our Premack principle article explores this in depth.

Common Implementation Mistakes

The first common mistake is under-reinforcing the alternative. The replacement behavior must pay more, more often, and faster than whatever was reinforcing the problem behavior. If the rate or value is too low, the animal stays with the original behavior. The second common mistake is allowing the problem behavior to continue to be reinforced in the background — by a family member who still greets the jumping dog, by an environment that still delivers the squirrel through the window, by the food bowl that still appears even when the dog has pushed the resident cat away. Every reinforcement channel for the problem behavior must be identified and closed.

A third common mistake is moving too fast. Differential reinforcement works through repetition, not insight. The animal does not “understand” the rule after one trial; the rule becomes installed through dozens of repetitions in clean conditions. Many handlers give up after three or four sessions and conclude that the procedure does not work, when in reality they have not yet given the procedure time to operate.

Frequently Asked Questions

Is differential reinforcement the same as ignoring the bad behavior? No. Ignoring is one possible consequence component, but differential reinforcement is fundamentally about what you reinforce, not what you withhold. The active piece is paying richly for the alternative behavior. Withholding reinforcement for the problem behavior is a secondary component, and it only works if you have correctly identified what was reinforcing the problem in the first place.

Can differential reinforcement be used for fear-based behaviors? Differential reinforcement on its own is not enough for fear-based behaviors, because the underlying emotional response is not changed by reinforcing an alternative operant behavior. Counter-conditioning and desensitization must be paired with it to change the dog or cat’s classical-conditioning history with the trigger. The combination is standard practice for reactivity, fear of strangers, and similar emotion-driven problems.

How long does differential reinforcement take to show results? Simple replacement behaviors in low-stakes contexts often shift in five to ten repetitions. Established problem behaviors in high-stakes contexts can take weeks to months of consistent application. The variable is how strong the historical reinforcement of the problem was and how cleanly the new contingency is implemented. Working with a credentialed behavior consultant accelerates the process dramatically.

Do I need a clicker for differential reinforcement? Not strictly, but a marker (clicker or verbal) sharpens the timing dramatically. The marker tells the animal exactly which behavior earned the reinforcer, which makes the learning faster and cleaner. A consistent verbal marker works adequately if the timing is clean.

When should I bring in a professional? If the problem behavior involves aggression, severe fear, resource guarding with bite history, or has not responded to a clean differential reinforcement protocol within four to six weeks of consistent application, bring in a credentialed behavior consultant or a board-certified veterinary behaviorist. Our veterinary behaviorist consultation cost article covers what to expect from the referral.