How long should a training session last?

A few minutes, repeated several times a day. Dogs learn better with short frequent sessions than with long lessons, because concentration drops quickly and tiredness brings mistakes that muddle learning. It is better to stop while the dog is still responding well, so each session ends on a success and the dog stays motivated for next time.

Do I have to use food treats?

Food is convenient because it is quick to deliver and most dogs find it motivating, but the principle is to reward with what the individual dog wants: for some a game or a stroke works just as well. What matters is that the reinforcement is immediate and that, when you use food, the treats are part of the daily ration so you do not add extra calories.

My dog obeys at home but not at the park: why?

Because the park is full of distractions, which are one of the three variables of stay and of behaviour in general. A cue learned in a quiet place does not transfer by itself to a busy one: it has to be retaught in the new context starting from an easy level, with low duration and distance. Generalising to different places is a phase in its own right, not a fault in the dog.

Should I push the dog into a sit with my hands if it will not?

No, avoid physically forcing it into position. Guide the movement with the treat over the head, so the dog chooses to sit, which makes learning more solid and free of conflict. If it does not respond, the movement is still unclear: make the hand cue smoother and try again, without repeating the word endlessly.

Training step by step10 minLevel: BeginnerUpdated on July 29, 2026

Teaching sit and stay step by step

Building sit and stay with positive reinforcement, splitting duration, distance and distraction, with clear criteria to move up or step back without forcing.

Audience: Pet owners
Species: Dog
Scope: Valid everywhere

What you need before you start

Premi piccoli e appetibili, tanti quanti bastano per una sessione, divisi dalla razione giornaliera così da non aumentare le calorie
Un ambiente tranquillo e povero di distrazioni per le prime sessioni, per esempio una stanza di casa
Sessioni brevissime di pochi minuti, ripetute più volte al giorno invece di una lunga
Una parola di rilascio scelta in anticipo, per esempio vai, che segnala la fine del resta

Sit and stay look like two trivial cues, and that is exactly why they are so often taught badly: too much is asked, too soon, in an environment full of distractions, and then people complain that the dog obeys in the kitchen but not at the park. They are in fact two different exercises. Sit is a single movement to pair with a word, and it comes together quickly. Stay is far harder, because it asks the dog to do nothing while something happens around it, and it has to be broken into separate variables: how long it stays put, how far you move away, how strong the distraction is. Anyone who mixes them all together fails. This guide first builds sit with positive reinforcement, then stay one variable at a time, with explicit criteria for moving up a level and for stepping back when the dog makes too many mistakes. No force: the dog learns because it pays off, not because it is compelled.

Why positive reinforcement works better

A dog repeats behaviours that bring it something good and drops those that produce nothing. Rewarding the right movement at the exact moment it happens makes that movement more likely next time, without the fear and conflict that come with coercive methods. The secret is not the amount of reward but the timing: the reinforcement has to arrive while the behaviour is happening, not a few seconds later, when the dog is already doing something else.

Many owners reward a moment too late and end up reinforcing the next movement, for instance the dog getting back up. That is why a signal marking the exact instant is useful, a short word that is always the same, said the moment the behaviour is correct and immediately followed by the reward. The signal tells the dog what it did well, the reward confirms it.

Building the sit

Guide the nose up and back
Hold a treat in front of the dog's nose and move it slowly up over the head, towards the tail. The muzzle follows the hand and the rear lowers naturally to the ground. Do not push the dog down with your hands: let the movement carry it into position.
Mark and reward the instant it sits
As soon as the rear touches the ground, say your marker and give the treat at once. Timing is everything here: if you reward while it is already getting up, you teach the wrong movement.
Repeat until the movement is fluid
A few repetitions per session, several sessions a day. When the dog sits easily following the hand, you are ready to add the word.
Add the word before the movement
Say sit once, calmly, just before guiding the movement. The word has to precede the movement, so the dog learns to associate it with the action it is about to perform.
Fade the hand cue
Gradually reduce the treat movement until the word alone is enough. If the dog does not respond, do not repeat the word several times: briefly go back to helping with the hand and try again.
Reward variably once it is solid
When the sit is reliable, move to rewarding not every time but unpredictably. Intermittent reinforcement makes the behaviour more stable, not weaker.

Stay broken into three variables

Stay asks the dog to hold a position while something changes around it. The difficulty depends on three independent variables, and the most common mistake is raising them all at once. You work on one variable at a time, keeping the other two low, and you raise the level only after several successes in a row.

The three variables of stay and how to grow them
Variable	What it increases	Rule
Duration	How long the dog stays still	Grows in small increments, before the steps back
Distance	How far you move away from the dog	One step at a time, always returning to reward
Distraction	What happens around while it stays	Introduced last, with duration and distance low

The stay procedure

Start from the sit and reward stillness
With the dog sitting, wait a single second, then mark and reward while it is still. At the start you do not even need the word stay: you are rewarding not moving.
Introduce the word and the release word
Say stay when the dog is still, then, after the reward, use your release word to tell it that it can move. The dog has to learn that the stay ends when you say so, not when it decides.
Extend duration in small steps
Increase the wait by a few moments at a time, irregularly, so the dog does not learn to get up at a fixed point. Keep distance and distraction at the minimum.
Add distance by bringing duration down
Take a single step back and return to reward at once, then two steps. Every time you raise distance, lower duration again: one variable at a time.
Introduce distractions calmly
Only when duration and distance are solid, add small distractions, mild at first then gradually greater, always bringing the other two variables back down.
Always release with your word
End every repetition with the release word, never letting the dog get up by itself. It is the release, not endless stillness, that defines a clean stay.
Close the session on a success
Finish while the dog is still responding well, not when it is tired and starting to make mistakes. A short session that ends well is worth more than a long one that ends in confusion.

Common mistakes and how to step back

Almost every sticking point comes from the same pattern: too much was asked and the dog stopped understanding. The cure is not to insist but to lower the level, rebuild a few easy successes and restart more slowly. A repeated mistake is not disobedience, it is a signal that the exercise needs to be made simpler.

The dog gets up at once: duration is too high, go back to rewarding after a single second.
It responds indoors but not outside: the environment is a distraction, restart from the base level in the new context.
You repeat the word several times: you teach that the cue only counts on the third time, say it once and help with the hand.
You reward late: you reinforce the wrong movement, work on the marker's timing.
The dog gets distracted halfway: you raised two variables together, bring one back to the minimum.
Sessions are too long: tiredness brings mistakes, better a few minutes several times a day.

Frequently asked questions

How long should a training session last?: A few minutes, repeated several times a day. Dogs learn better with short frequent sessions than with long lessons, because concentration drops quickly and tiredness brings mistakes that muddle learning. It is better to stop while the dog is still responding well, so each session ends on a success and the dog stays motivated for next time.
Do I have to use food treats?: Food is convenient because it is quick to deliver and most dogs find it motivating, but the principle is to reward with what the individual dog wants: for some a game or a stroke works just as well. What matters is that the reinforcement is immediate and that, when you use food, the treats are part of the daily ration so you do not add extra calories.
My dog obeys at home but not at the park: why?: Because the park is full of distractions, which are one of the three variables of stay and of behaviour in general. A cue learned in a quiet place does not transfer by itself to a busy one: it has to be retaught in the new context starting from an easy level, with low duration and distance. Generalising to different places is a phase in its own right, not a fault in the dog.
Should I push the dog into a sit with my hands if it will not?: No, avoid physically forcing it into position. Guide the movement with the treat over the head, so the dog chooses to sit, which makes learning more solid and free of conflict. If it does not respond, the movement is still unclear: make the hand cue smoother and try again, without repeating the word endlessly.

What to do next

Build the sit first by guiding the nose up and marking the instant the rear touches the ground, then add the word. For stay, work on one variable at a time, duration, distance and finally distraction, keeping the other two low and always closing with the release word. Sessions of a few minutes several times a day, and at every repeated mistake lower the level instead of insisting.