How to Teach Your Dog "Leave It"

A "leave it" that works on dropped food, dead things on the sidewalk, and another dog across the street — built from a simple exercise your dog can learn today.

If your dog's nose hits something on the ground and their ears turn off, you're not alone — and you're not dealing with a stubborn dog. "Leave it" is one of the most important safety commands, and the reason it falls apart in the real world is almost always that it was never built for the real world. The steps below start simple and scale to the situations that actually matter. When something goes wrong, we tell you exactly why and what to adjust.

Before you start

Find a quiet room. Have two types of treats: something decent in one hand (this is the "bait" your dog will practice ignoring) and something your dog loves in the other hand (this is the reward for ignoring it). The reward must be better than the bait — otherwise you're asking your dog to give up something good for something worse, and that's a negotiation you'll lose.

If your dog can't focus on you right now — pacing, whining, fixated on something in the room — the environment is too stimulating for learning. Try a smaller, quieter space. A bathroom with the door closed works. That's not overkill — it's setting conditions where your dog can actually think.

Why some dogs can't focus in certain environments →

The Steps

Step 1: The closed fist

Hold a treat in a closed fist and present it to your dog at nose level. Let them do whatever they're going to do — paw, lick, nudge, nibble at your fingers. Say nothing. Wait.

If that worked: The moment your dog backs off from your fist — pulls their nose away, looks away, or even hesitates — say "yes" and reward with a different treat from your other hand. Not the treat in the closed fist. A different, better one.

This distinction matters. The dog learns: ignoring that thing gets me something better from somewhere else. The treat they left alone is not the treat they earn.

Your dog won't stop pawing and licking at your fist?

Good — this means your dog is an operant dog, actively experimenting with strategies to get what they want. That's the raw material you need. They just haven't tried the one strategy that works yet: backing off.

Try this: Keep your fist closed and wait. Do not pull your hand away. Do not say "no." Just be a statue. Most dogs will eventually get frustrated and pull back — even for a split second. The instant you see any disengagement — a pulled-back nose, a glance away, a tiny pause — say "yes" and reward from the other hand. You're capturing the direction, not the final behavior. The duration of the "leave it" comes later.

If your dog has been working your fist for more than 30 seconds straight without any pause at all, close your fist inside both hands to make it completely inaccessible. Wait. The moment of disengagement will come.

Your dog sniffs once and then walks away, uninterested?

The bait treat isn't interesting enough to create the conflict you need. If there's nothing worth ignoring, there's nothing to learn from.

Try this: Use a higher-value bait treat — something that your dog will actually work for. You need them to want the thing in your fist. Then the moment they choose to disengage from it, that choice is meaningful and worth marking. Your reward treat still needs to be better than the bait. If the bait is cheese, the reward might be chicken.

Step 2: Mark the disengagement

Repeat Step 1 until your dog is regularly backing off from your closed fist within a few seconds. You're looking for a dog who sniffs your fist, hesitates, and then looks at you or looks away — that pause is the behavior you're building.

If that's happening: Your dog is starting to understand the game: ignoring the thing in the fist is what makes the better thing appear. Repeat 8-10 times until this pattern is smooth.

Your dog backs off but then immediately lunges back at your fist?

Your timing on the marker might be slightly late. The dog is backing off, but if "yes" doesn't land in that split second of disengagement, they re-engage before the reward arrives. Dogs live in a 1-to-3-second window — if the mark doesn't capture the right moment, they don't know what earned the reward.

Try this: Speed up your "yes." Say it the instant you see the nose pull back — not when the dog has fully turned away, not when they make eye contact. The very first frame of disengagement. If you're consistently late, try watching your dog's nose specifically. The moment it moves away from your fist, even a fraction of an inch, mark it.

Quick check: Is your dog engaging in this game eagerly? Are they sniffing the fist, then looking at you, then getting rewarded — and repeating the loop with enthusiasm? If they're yawning, lying down, sniffing the floor, or just staring blankly, they need a break. Most dogs can do 3-5 minutes of focused training at a time. End on a success and come back later.

Step 3: Add the words

Once your dog is reliably disengaging from your closed fist (you can predict they'll back off within a few seconds), add the verbal cue. Say "leave it" just before you present your fist. One time. Clear and calm. Not loud, not stern.

If that worked: Your dog hears "leave it," sees the fist, backs off, you say "yes," reward from the other hand. Repeat until "leave it" + fist presentation smoothly produces the disengagement.

Your dog was doing great but seems confused now that you've added the words?

Adding a verbal cue is a new piece of information. Some dogs briefly stall when the picture changes, even slightly. This is normal and temporary.

Try this: Go back to a few reps without the word to rebuild momentum. Then add "leave it" again, but say it casually — almost as an afterthought before presenting the fist. The fist is still the main signal. The word will absorb meaning through repetition. Don't force it. Over the next few sessions, the word starts carrying more weight on its own.

Step 4: Treat on the floor

Place a treat on the floor and cover it with your hand. Say "leave it." When the dog disengages, mark and reward from the other hand. Gradually lift your hand off the treat — an inch at a time — until the treat is sitting on the floor uncovered and your dog is choosing to ignore it.

If that worked: Your dog is looking at a treat on the ground and choosing not to take it. This is real impulse control being built in real time.

Your dog dives for the treat the moment you move your hand?

The uncovered treat on the floor is a much bigger temptation than a treat in your hand. The dog can see it, smell it at full strength, and there's nothing between them and it. You've jumped up in difficulty, which means you may need to slow down.

Try this: Move your hand away in smaller increments. Lift one finger. If the dog holds, mark and reward. Then lift two fingers. Then raise your hand half an inch. Then an inch. Each increment is its own success. If the dog breaks at any point, cover the treat again — no scolding — and back up to the last level that worked. This is tedious. It also works.

Your dog leaves the floor treat but then slowly creeps toward it?

Your dog is testing the boundary — not being defiant, just seeing where the edges are. The initial disengagement was genuine, but without a quick reward to capture it, the temptation pulls them back.

Try this: Mark and reward the instant they disengage, before the creep starts. If you see them hold still for even one second with the treat uncovered, say "yes" and reward from your other hand immediately. Speed of reinforcement is everything here. Once the pattern is solid at one second, you can start building to two, then three. Don't let the creep develop — capture the good decision before the bad one starts.

Step 5: Add movement

Drop a treat from waist height. Say "leave it" as it falls. The moving treat is significantly more exciting than a stationary one — it triggers drive. Be ready to cover it with your foot if the dog goes for it.

If that worked: Your dog watches the treat fall, holds back, looks at you. Mark and reward generously. This is a big step — moving objects are much harder to resist than still ones.

Your dog ignores stationary treats perfectly but loses it when the treat moves?

Movement activates a completely different part of your dog's brain. The instinct to chase or snatch a moving thing is drive — it's hardwired, not learned. A stationary treat tests impulse control. A moving treat tests impulse control against instinct. These are not the same level of difficulty.

Try this: Bridge the gap. Start by moving the treat very slowly in your hand — just wiggling it — while saying "leave it." Mark the disengagement. Then move it a bit more. Then roll it on the floor slowly. Then drop it from 6 inches. Then a foot. Then waist height. You're building a ladder from "still treat" to "moving treat" with small enough steps that the dog succeeds at each one.

If this isn't clicking yet

Some dogs get this in a single session. Some need a week of short daily practice. Both are normal, and the speed tells you nothing about your dog's intelligence or your skill as a trainer.

If you've tried the adjustments above and you're still stuck, the issue is almost always one of these three things:

  1. The environment is too hard. Strip it back further than you think you need to. A quiet room with no other people or animals and nothing interesting on the floor. If your dog can't focus on the fist exercise here, something else is going on — anxiety, overstimulation, or a need to burn energy first. A walk before training can change everything.
  2. The reward isn't competing. Your "leave it" reward needs to be better than the thing you're asking the dog to leave. If the bait and the reward are the same value, the dog has no reason to disengage. The equation must be clear: ignoring that thing gets me something better. Upgrade your reward treat.
  3. Your dog needs more guidance. Some dogs naturally experiment and figure out that backing off works. Others stare at your fist and wait for instruction. If your dog is in the second camp, you can help: when you present the fist, move it slightly to one side. When the dog follows it and then resets, they may naturally disengage for a moment. Capture that moment. You're creating the opportunity for the behavior to happen, then marking it.

These aren't character flaws in your dog. They're variables. Adjust the variables.

Making it harder (gradually)

Once your dog reliably leaves treats on the floor in your training spot, you can start changing one variable at a time:

The single biggest mistake in this phase is changing too many things at once. If you're practicing on a walk for the first time, that IS the variable. Don't also try it with a higher-value item at a greater distance. One variable. Your dog will tell you if they can handle it.

A suggested progression for real-world "leave it":

  1. Treats on the floor in your quiet training room
  2. Treats on the floor in the kitchen (new environment)
  3. You drop a treat in the kitchen while your dog is nearby
  4. Food on a low table or counter edge (only if safe to set up)
  5. Treats placed on the ground in the backyard
  6. Low-value items on the sidewalk during a calm walk
  7. Higher-value distractions on the sidewalk (food wrappers, other dogs at a distance)

If the dog fails at any stage, you moved too fast. Go back to the last point where they succeeded and stay there longer. That's not a setback — it's the dog telling you where their current limit is.

Important — management between training sessions: Until your dog's "leave it" is reliable in a given environment, manage the situation so they can't practice the wrong behavior. On walks, keep them on a short enough leash that they can't reach things on the ground before you can intervene. In the house, don't leave tempting food at nose level. This is not a lack of trust — it is management, preventing the dog from being rewarded for the exact behavior you're trying to eliminate. Every time the dog grabs something they shouldn't and it tastes great, that behavior gets stronger.

Why this works

Every progression step on this page is a test of your management as much as your dog's impulse control. If the dog fails at stage 5, you moved too fast — and "too fast" is a handler decision, not a dog decision. Your half determines the outcome.

When your dog backed off from the treat and you said "yes," you marked the precise moment they chose to disengage from something they wanted. This is operant conditioning at its clearest: the dog tried a strategy, discovered it worked, and started choosing it on purpose. Rewarding from the other hand taught a second lesson — the good thing doesn't come from the thing you're leaving, it comes from somewhere else entirely.

How Learning Works → — the full picture of how dogs form connections between choices and outcomes The Tools → — markers, luring, and why the timing of "yes" matters so much

Go deeper