Loose Leash Walking
It's important to teach dogs without causing them pain. So, by experimentation,
experience and help, we've come up with a few gentle methods to teach a dog NOT to
pull on a leash.

Before beginning, we must think about what a leash means to an untutored dog: a rope
tied to her collar which allows her to pull her owner wherever she - dog - wants to go.
We, of course, would like the dog to see the leash as a rope that allows us to keep her
by our sides. Unfortunately, that takes some doing. So here are a few methods to try.
At least one should suit your training style and your dog's learning style.
With any method, it's extremely important to remember that the dog needs to learn to
look at you. You cannot force that; it must come because you are being unpredictable.
I call it being predictably unpredictable, because the dog knows you are likely to
change directions or even disappear without much notice, and it's her job to make sure
you don't.

METHOD ONE
Attention Walk
Begin with the dog sitting directly in front of you; treats in your hand, leash very
loosely held or tied to your waist.
Take a step or two backwards, the dog follows. You stop, she sits. You say "Yes"
(marker word for correct behavior) and give a treat.
Do this several times
Now, begin to extend the behavior, walking back for several feet before you stop and
reward.
After she's got this (it takes maybe 4 minutes!), Begin to teach her to walk beside you
by beginning the behavior - walking backwards, then abruptly walking INTO the dog,
pushing her gently to your side. Take three for four steps, stop, she sits beside you,
rather than in front and you say "Yes!" and treat.
Extend the behavior, each time making her go further and further beside you. If she
begins to forge ahead, quickly step backwards, she follows, move into her again, stop,
sit, YES!
Later, when the behavior begins to be reliable, add the cue. As you begin walking, say
"Let’s Go," thus telling her what's she is doing right!

METHOD TWO
Target Walking
Begin to walk with the dog towards a delicious target - a bowl of chicken beef, for
instance, that the dog knows is there. As soon as she steps ahead of you, say "oops,"
stop, and walk back to the start line, pulling the dog along with you. Begin again. The
dog gets the treats when she has successfully walked WITHOUT PULLING all the way
to the bowl.

METHOD THREE
We aren't going that way!
This method does not use treats, which makes it very appealing to some. It's very easy
to make this a painful procedure, if the dog has the wrong kind of collar, or you end up
jerking the dog back to you. Use a flat, rolled or martingale collar for this exercise.
Hold the dog's leash very loosely, giving her about three to four feet of slack. Your
hands should be held tightly against your chest to avoid giving the dog any cues. You
want her to watch you.
Begin walking, with her by your side.
The instant she begins to step ahead, say "oops" or "uh-uh" (again, mild but
consistent!), and take off in another direction - any direction except the one you were
going in. When she catches up, praise her and we keep going.
Repeat as necessary. Add the cue when the behavior becomes reliable.

METHOD FOUR
The Double Leash Method
I usually begin with a young puppy, but this can be done with an older dog, as long as
you apply it consistently. It's great for small dogs.
There are three essential ingredients. Lots of soft, easily broken up treats, a regular 6-
foot leash and a long line (20-30 feet). The line should be light, but does not have to be
an "Official" long-line. Nylon cording works just fine, especially with knots every three
feet or so.
Begin by teaching the dog that being close to you is always rewarding. When she comes
to you voluntarily, she gets praise, pets, attention and a treat or two. You might drop
several small treats on the ground, to make the area around you even more attractive.
But try not to be predictable about where the treats come from - sometimes from your
hand (when she sits), other times from the ground.
Now, put the long-line on, but not the regular leash. She should get used to dragging
the line; sometimes dogs find it disconcerting for awhile. We want her to disregard it.
This usually takes a few minutes.
As you begin to walk, she'll probably go ahead of you (it's the "Doggy Way"). When she
gets to a point beyond which you do not want her to be, give her one gentle word -
could be "wait" or "stop"…doesn't really matter, as long as it's the same one.
Immediately after you've said the word, step on the trailing leash. She abruptly halts;
say her name and begin walking in another direction - not necessarily back the way you
came. Most dogs very quickly catch up and begin walking by your side, at which point
she is rewarded. Continue the walk. When she pulls ahead, say "stop," and step on the
leash. When she comes back, praise and begin walking again. Many, many times.
After she's keeping an eye on you all the time, add the normal leash; hold it VERY
LIGHTLY and quite loosely. It should be loose enough to slip through your hand if the
dog pulls and you don't step on the trailing leash. Do the same exercise as before, using
the stop or wait cue, and NOT PULLING ON THE REGULAR LEASH AT ALL. That
leash should be there for looks and the law only, not for correcting. This is very
important, since dogs easily learn that pulling the leash forces you to go with them. I
don't want my dog to learn that.
Go through the same procedure you did before (by now the word "stop" or "wait" is
enough to stop the dog). As you turn and walk in the other direction, and she walks by
you, cue her with the words "walk with me" (or - if you wish "heel.") Twist and turn and
go in many different directions, often stopping and rewarding. Always keep your hands
so loosely on the leash that she gets NO cues from the leash, only from watching your
body.
Try to remember always - You're the leader - she's the follower. It's the only way it'll
work!
No matter what method or methods you choose, they all take patience, persistence,
practice….and praise!

T. King, September, 2000