Dog pulls on the leash and you've tried everything: what's missing, and what works

When a dog still pulls after every tip, tool and video, the usual reason is simple: pulling still works some of the time. The dog moves forward on a tight leash often enough to keep trying. The San Francisco SPCA puts it plainly: if you allow even a few steps of pulling before you stop, you are rewarding the pulling.

Three other things keep it going. New gear that makes pulling harder but never teaches the dog what to do instead. Training on the busiest street at the busiest time. And pulling that isn't about walking at all: a dog that hauls toward other dogs, people or bikes is often a reactive dog, and that needs a different plan.

What works is not one more trick. It is one rule kept on every walk (a tight leash never moves forward), food paid at your side, short practice where it's easy, and a front-clip harness to manage the rest of the time.

Pulls toward dogs or people? Take the free check

Why pulling survives "everything"

Pulling still pays off. Every time a dog pulls and gets to the tree, the smell or the park, pulling is rewarded. The RSPCA says the stop must happen every single time the dog moves away, so the dog learns that not pulling is the fastest way to get where it wants to go.
The rule changes from walk to walk. One person stops on a tight leash, another keeps walking when late for work. To the dog, pulling works sometimes, and behavior that pays off sometimes is hard to stop.
The gear held it down but didn't teach. A harness, halter or new collar changes how hard it is to pull. The RSPCA in Australia calls even the front-attach harness a management tool. Without the training underneath, the pulling is still there.
Your pace isn't theirs. The SF SPCA notes that a dog's normal walking speed is faster than yours, which alone makes it hard not to pull.
The leash sets it up. The RSPCA in Australia says a short leash sets a dog up to pull, and retractable leashes encourage pulling. The SF SPCA says a retractable leash gives to a pulling dog and teaches it to pull.
Practice happens in the hardest place. Most dogs only "train" on real walks, full of smells, squirrels and other dogs. Both the SF SPCA and the RSPCA start in a quiet place with nothing to chase.
Too much at once. A dog that has been inside all day, or that is excited or worried on the walk, finds it hard to focus on a slack leash. Training is easier when the dog is calm enough to take food.

What owners try, and what to change

What owners try	Why it often stalls	What to change
Stopping when the dog pulls	Done some walks, not others; or the stop comes after a few steps	Stop the moment the leash goes tight, on every walk, with everyone who walks the dog
Treats	Given in front of the dog, or only when it comes back	Pay often, beside your leg, while the leash is loose
Turning around	Used as a jerk on the leash	Turn calmly, then reward the dog for catching up on a loose leash
A new harness or head halter	Used instead of training	Use it to manage walks while you train in short sessions
Leash pops, choke or prong collar	Can stop pulling for a moment; carries a risk of pain, fear and aggression (below)	Drop them; reward-based methods do as well or better
Long walks to tire the dog out	Long walks on a tight leash are long practice at pulling	Short training sessions, plus separate sniff time where pulling isn't possible

What works, step by step

This is the method shelters such as the SF SPCA and the RSPCA teach, put in order.

Set up. A flat collar or a well-fitted harness, and a fixed leash of about six feet (the SF SPCA) or about 2 meters (the RSPCA in Australia), not a retractable one. Small, soft treats your dog loves. A quiet place: a yard, an empty parking lot, a hallway.
Red light, green light. Start walking. The moment the leash goes tight, stop and stand still, "like a tree," in the RSPCA's words. Wait. When your dog turns back and the leash goes slack, praise and walk on. The SF SPCA's advice: watch the leash, not the dog. It doesn't matter why the leash went loose, only that loose means you walk and tight means you stop.
Pay the position, not the return. Feed from the hand on the side your dog walks, right beside your leg, every few steps while the leash is loose. The RSPCA in Australia says to reward often and with something your dog really likes, every few steps at first. Food at your hip builds the habit of staying there.
Change direction. When your dog surges ahead, turn and walk the other way, calmly and without a jerk. When your dog catches up on a loose leash, reward at your side. A practiced U-turn is also how you get out of trouble later.
Keep sessions short. The SF SPCA suggests about 10 minutes, two or three times a day, until your dog is good at it. Stop while it's going well.
Add distractions slowly. The RSPCA's exercise: put a toy on the floor. If your dog pulls toward it, stop and call your dog back. Only a loose leash gets your dog to the toy, so your dog learns that pulling slows things down. Then a quiet street, then a busier one.
Manage every other walk. Until the skill is solid, real walks will undo practice if pulling works on them. Use a front-clip harness (below), pick quieter routes and times, and keep the stop rule as much as you can.
Give sniffing its own time. Many dogs pull because they want to explore. Give them time to do it on a long line in a safe, open space, or in a "sniff walk" where you let your dog choose the pace. In a 2019 study by Duranton and Horowitz, dogs that did nosework daily for two weeks became more "optimistic" in a judgment test; dogs that did heelwork instead didn't change.

If the stop rule feels too slow on a real walk, that is normal at first. The RSPCA says loose-leash walking takes time, patience and consistency.

Is it pulling, or is it reactivity?

Look at when your dog pulls. A dog that pulls all the way to the park, at a steady pace, is pulling to get somewhere. A dog that walks fine until it sees another dog, a person, a bike or a car, then hauls toward it, stares, barks or lunges, is reacting to something. Stop-and-go won't fix that, because the dog isn't thinking about the leash.

For a reactive dog, the work is about distance: find how far away the trigger can be while your dog still notices it calmly, train just beyond that line, and close the gap slowly. Our guides go through it: why dogs lunge at other dogs on leash, barking at other dogs on walks, and reactive dog training at home.

The free dog check asks 10 questions about what your dog does on walks, finds your dog's trigger distance and gives you the first 3 days of drills.

Harnesses, halters and prong collars: what the research says

Front-clip harness

A front-clip (front-attach) harness has the leash ring on the chest, so a pulling dog is turned to the side. In a 2024 study by Johnson and Wynne, 23 dogs were walked for 5 minutes each on a martingale collar, a front-connection harness, a plastic prong-style collar and a metal prong collar, with a gauge on the leash. The harness and both prong collars all reduced pulling compared with the martingale, and there was no significant difference between the harness and the prongs. In other words, the prong collar did not beat the harness. The authors found no clear evidence of poorer welfare in those short walks and caution against generalizing from them.

The RSPCA in Australia calls a front-attach harness the most appropriate management tool for most dogs that still pull, and suggests a double-ended leash: one end on the chest ring, one on the back ring or collar.

Head halter

A head halter steers the head, and the body follows. The SF SPCA suggests one for walks while you train. The RSPCA in Australia is more cautious: many dogs find head collars uncomfortable, a jerk can turn the head sharply, and dogs need to be taught to accept them, so it advises trying loose-leash training and a front-attach harness first. If you use one, introduce it slowly, with food, and never jerk the leash.

Collars and the neck

In a 2020 study by Carter, McNally and Roshier, seven collars and a slip lead were tested on a model of a dog's neck. None gave a pressure low enough to rule out a risk of injury when the dog pulls.
In a 2006 study by Pauli and colleagues of 26 dogs, pressure in the eye rose when dogs pulled against a collar but not against a harness. The authors advise a harness for dogs with glaucoma, thin corneas or other conditions where higher eye pressure could harm.

Prong, choke and shock collars

The American Veterinary Society of Animal Behavior's position statement on humane dog training (2021) says to avoid choke chains, prong collars, shock collars and leash jerking. It lists the fallout as more anxiety, fear-related aggression and avoidance, and finds no evidence that aversive methods work better than rewards. The RSPCA in Australia says check chains and prong collars must not be used, because they cause pain and distress and can damage the neck. A 2017 review by Ziv found no evidence that punishment works better than reward-based training, and some evidence that it works worse.

For a dog that pulls toward other dogs, there is one more risk. The SF SPCA warns that choke, pinch and shock collars cause pain, which does not reduce frustration and can make aggression worse.

When to see a vet or a trainer

A sudden change. A dog that started pulling, lagging or reacting out of the blue, or seems stiff, sore or touchy, should see a vet first. A 2020 paper by Mills and colleagues estimated that around a third of dogs referred for behavior problems had a painful condition.
Eye, neck or breathing problems. Ask your vet about a harness instead of a collar (see Pauli above).
Still pulling after weeks of steady practice. The RSPCA in Australia advises a trainer who practices reward-based training. Ask how they train before you book; a trainer who reaches for a prong or shock collar is not following AVSAB's advice.
Bites. If your dog has bitten a person or another dog, see a board-certified veterinary behaviorist before you start any home plan. Find one in the specialists' directory.

How long it takes

It depends on the dog, how long the pulling has been paying off, and how consistent the walks are. None of the sources give a fixed time. A dog that has pulled for years has had years of practice, so expect weeks of short sessions, not a weekend. Count progress in what you can measure: how many stops a 10-minute session needed this week compared with last week.

Questions owners ask

Should I just use a prong collar if nothing else works?

In the one study that compared them, a front-clip harness reduced pulling about as much as a prong collar. AVSAB and the RSPCA both advise against prongs. Change the training first.

Will a harness make my dog pull harder?

A harness takes the pressure off the neck. A front-clip harness also turns the dog to the side when it pulls; the 2024 study found it reduced pulling. Use it with the stop rule, not instead of it.

Should I tire my dog out before a walk?

Some play or a sniff in the yard can help a dog that has been inside all day to focus. But a long walk on a tight leash is long practice at pulling. Keep training short and separate from exercise.

Can my dog still sniff?

Yes. The SF SPCA's goal is a dog that can explore and sniff on a six-foot leash without pulling. Let sniffing be a reward: a loose leash earns "go sniff."

My dog only pulls toward other dogs. Is that the same problem?

Usually not. That is often leash reactivity, and it needs distance work. Read why dogs lunge at other dogs, then take the free check to find your dog's trigger distance.

This is the method our plan follows, walk by walk: see how the plan works, or look at a sample day.

Find your dog's trigger distance (free check)

Sources

San Francisco SPCA. Dog: loose-leash walking (handout). sfspca.org
RSPCA. How to teach your dog to walk nicely on a lead. rspca.org.uk
RSPCA Australia. What equipment should I use when teaching my dog or puppy to walk on a leash? Updated January 10, 2023. kb.rspca.org.au
Johnson AC, Wynne CDL. Comparing efficacy in reducing pulling and welfare impacts of four types of leash walking equipment. PeerJ 12 (2024): e18131. doi.org/10.7717/peerj.18131
Carter A, McNally D, Roshier A. Canine collars: an investigation of collar type and the forces applied to a simulated neck model. Veterinary Record 187 (2020). doi.org/10.1136/vr.105681
Pauli AM, Bentley E, Diehl KA, Miller PE. Effects of the application of neck pressure by a collar or harness on intraocular pressure in dogs. Journal of the American Animal Hospital Association 42 (2006): 207–211. doi.org/10.5326/0420207
American Veterinary Society of Animal Behavior. Position statement on humane dog training, 2021. avsab.org
Ziv G. The effects of using aversive training methods in dogs: a review. Journal of Veterinary Behavior 19 (2017): 50–60. doi.org/10.1016/j.jveb.2017.02.004
Duranton C, Horowitz A. Let me sniff! Nosework induces positive judgment bias in pet dogs. Applied Animal Behaviour Science 211 (2019): 61–66. doi.org/10.1016/j.applanim.2018.12.009
Mills DS and colleagues. Pain and problem behavior in cats and dogs. Animals 10(2) (2020): 318. doi.org/10.3390/ani10020318
American College of Veterinary Behaviorists. Find a board-certified veterinary behaviorist. dacvb.org