The same treat that earns you a perfect recall in the back garden can be completely ignored the moment a squirrel, another dog, or an unfamiliar smell enters the picture. This is the single most useful thing to understand about training rewards: value is decided by your dog, in that setting, on that day, and it changes from one minute to the next.
This article explains how to choose rewards that actually work, how to match them to the difficulty of what you are asking, and how to manage the food side of it without quietly making your dog overweight. It is one of our training guides and sits alongside the pieces on reward markers and timing and why your dog listens at home but not outside. It is for owners who have tried training with treats and watched their dog's interest vanish at exactly the wrong moment.
The dog decides what counts as a reward
A reward is anything your dog will work to earn, and only your dog gets a vote. The Dogs Trust is explicit that all dogs are different and will find different things rewarding, while most enjoy food, toys, and our attention. A squeaky toy that one dog would climb over you for leaves another dog cold. Cheese that one dog treats as treasure is just food to the next.
This means your first job is observation, not shopping. Watch what your dog actually pursues. Do they push into your hand for food, grab a toy and parade it, or light up when you reach down to fuss them? What do they choose when nothing is on offer and no one is watching? That is your reward list.
Food is the practical default because it is fast, repeatable, and almost always interesting. But it is a default, not a law. A dog who would genuinely rather play tug than eat has just told you their high-value reward, and you would be leaving the best tool in the drawer unused.
Testing your dog's preferences without buying the whole shop
You do not need a scientific preference test, and you do not need to buy twenty varieties of treat. You need a handful of options and a few quiet minutes. Lay out two or three candidates, such as a piece of dry biscuit, a small piece of chicken, and a toy, and notice which your dog reaches for first when all three are freely available.
A slightly better test is to put the options in a place where the dog has to choose between them at the same time, rather than one after another, because dogs, like people, will often take the first thing offered and you learn nothing. Side by side, the preference shows itself in seconds.
Do this a few times over a week, because preference is not fixed. A dog who ignores a toy on a lazy afternoon may choose it over food after a rest, and the dog who loves chicken today may be indifferent tomorrow. The goal is a rough map of what your dog usually values, not a permanent ranking chiselled in stone.
Matching the reward to the difficulty of the task
The core idea is simple and is repeated across welfare sources. Use modest rewards for easy things, and save the exciting ones for hard things. The Dogs Trust suggests using less exciting treats such as dry dog biscuits or chopped veg to reward your dog at times it is easy for them to get it right, and reserving soft, smelly treats such as cheese, sausage, or chicken for the gold-star moments and for busier environments.
The reasoning is practical, not stingy. If you hand out chicken for a sit in an empty kitchen, you have nothing left to offer when a cyclist goes past and you need your dog's attention. Keeping a range means you can always trade up when the challenge climbs.
A useful way to hold this in your head is to ask, what am I competing with right now? In a quiet room you are competing with almost nothing, so a piece of their own kibble may be plenty. On a pavement you are competing with smells, movement, and other dogs, so the reward needs to be strong enough to matter more than all of it.
Building a simple reward ladder for your own dog
You do not need a formal system, but it helps to sort your dog's rewards into two or three rough tiers. The exact names do not matter. What matters is that you know, before a session, what is ordinary and what is special.
The everyday tier is for known behaviours in calm places: their normal food, a piece of dry biscuit, or a calm stroke. The strong tier is for new behaviours, more distracting settings, and anything that needs extra effort: soft, smelly, high-value food, or a favourite toy. A rare jackpot is for the biggest asks, such as coming back when called away from another dog.
The Dogs Trust describes this directly: soft and smelly treats such as cheese, sausage, or chicken work as the gold-star reward when your dog gets something spot on, and as the thing that keeps their focus where there are more distractions. Chopped into pea-sized pieces, they go a long way.
The three basic shapes rewards come in
Food is the workhorse. It is quick to deliver, easy to repeat, and can be varied in value. The main skill with food is portioning it small and counting it toward the day's allowance, which the next section covers.
Toys and play suit dogs who are more motivated by a game than a mouthful. A quick tug or a thrown ball can be a powerful reward, and the Dogs Trust notes that you can use fetch or tug to reward good behaviour once you have taught the dog to release the toy on cue. Play also tends to raise energy, which can be useful or unhelpful depending on what you are training.
Attention is the quiet giant. The Dogs Trust points out that most dogs feel good any time we look at, touch, or talk to them, which means we are constantly influencing behaviour with our attention. A calm word or a stroke rewards a settled dog without winding them up, and for some dogs attention is the most valuable currency you own. The catch is that attention is also the easiest reward to spend by accident, rewarding jumping or barking simply by reacting.
Size and texture matter more than the label
A reward that takes twenty seconds to chew brings the session to a halt. The best training food is small and fast to eat. Blue Cross suggests a treat about the size of your little fingernail for dogs, and the Dogs Trust recommends chopping treats into pea-sized pieces. Tiny pieces let you deliver many rewards in a short session without your dog filling up or losing momentum.
Soft food usually works best for fast repetition because it can be swallowed quickly. Hard or crumbly treats can still work if they are small, but if they crumble into dust in your pocket they are better saved for something slower. The label on the packet matters far less than whether you can deliver it cleanly and quickly, ten times in a row.
Wet, aromatic food tends to be higher value than dry because smell carries it. A small piece of plain cooked chicken or a smear of something soft will almost always outrank a dry biscuit in a distracting place, which is precisely when you need that edge.
Keeping rewards safe, within budget, and non-toxic
Training treats are food, and they add up faster than they feel. PDSA recommends making sure treats do not take up more than ten per cent of your dog's daily food allowance, with their usual food reduced by the same amount so they are not overeating. A practical version is to set aside a portion of the day's food as the treat allowance before you start.
The Dogs Trust makes the same point from the other direction: if you use food for training, cut down on meals to prevent weight gain. This is not about denying your dog, it is about accounting, because a handful of cheese delivered one small piece at a time is still a handful of cheese.
Whatever you choose must also be safe for your dog. Some human foods are toxic to dogs, and a treat that suits one dog may upset another's stomach. When in doubt, use food you know your dog tolerates, keep pieces plain, and check anything unfamiliar before it goes in the pouch.
A few everyday foods are genuinely dangerous to dogs, and the list is worth knowing cold: chocolate, grapes and raisins, onions and garlic, and anything containing the sweetener xylitol are all ones to keep out of the training pouch entirely. This is not an exhaustive list, and if you are unsure about a particular food, ask your vet rather than guessing.
When you do use human food, keep it plain and unseasoned. A small piece of plain cooked chicken is a superb training reward; chicken roasted with onion, garlic, or salt is not. The simpler the food, the fewer things can go wrong.
If your dog has a medical condition or is on a prescription diet, check with your vet before introducing new treats, because even a healthy food can be the wrong choice for a dog with specific needs. For a dog with a sensitive stomach, introduce any new reward one at a time so you can spot a reaction.
Life rewards you are already carrying
Some of the strongest rewards are things your dog wanted to do anyway, and they cost you nothing. Getting to sniff, running off lead, being released to greet a friend, or the door opening are all real rewards for many dogs, and you can use them deliberately.
The Dogs Trust describes this as using rewards within the environment. If your dog loves running off lead, that freedom can be the reward for walking nicely for a few paces. If they love meeting people, getting to say hello can reward keeping four paws on the ground. The reward is not something you pull from a pocket; it is permission to do the thing they already want. This is the same idea behind our guide to loose-lead walking, where the sniff or the freedom is the currency.
This is powerful because it teaches your dog that good choices unlock the things they care about, but it only works if you are consistent about the exchange. If the sniff arrives whether or not they walked nicely, it was not a reward, it was just Tuesday.
Reading when your reward has stopped working
A dog who normally loves chicken but turns away from it on a busy street is not being fussy. They are over threshold, too worked up to take food, and that is information about the environment rather than the treat. In that moment the fix is usually distance or a quieter spot, not a bigger piece of chicken.
A dog who takes the treat but immediately re-engages with whatever they were doing has been paid, not distracted. The reward worked, but the surrounding thing was more compelling, which means the difficulty was set too high for the reward you offered. Trade up or move somewhere easier.
A dog who refuses food altogether during training, even somewhere calm, might simply not be hungry, or food may not be their currency that day. Try a toy or attention instead before concluding that training is the problem. Rewards are perishable and personal, and a dead reward is often just a wrong guess, not a broken dog.
Rotating rewards so nothing goes stale
Even a brilliant reward loses its shine if it is the only thing on the menu for weeks. Rotation is not about being fancy, it is about keeping the reward valuable by keeping it a little scarce. If chicken appears at every single session, chicken becomes ordinary, and the thing you were saving for hard moments quietly stops being special.
A simple approach is to keep two or three rewards in rotation and swap them across sessions, while holding the highest-value one back for the hardest work. The everyday tier can rotate freely; the jackpot should stay rare enough that it still means something when it appears.
Rotation also protects you against the day when your dog is simply bored of the usual thing. Having a second and third option means a flat session can be rescued by a change of reward rather than abandoned as a failure.
The moment the reward arrives matters
A reward delivered slowly teaches a different thing than one delivered instantly. The dog associates whatever they were doing when the reward landed, so a delay means you may be rewarding the pause, the glance away, or the movement that came after the good behaviour.
This is where a marker earns its keep. A marker is a short signal that tells your dog the exact moment they got it right, and it is always followed by the reward. Pairing your reward with a marker lets you be precise even when your hand is slow. Our guide to reward markers and timing covers how to build that loop.
For now the rule is simple: have the reward ready before you ask, and deliver it the instant the behaviour happens. Fumbling in a pouch after the fact is how good intentions turn into rewarding the wrong moment.
Rewards for a puppy versus an adult dog
Puppies are learning what training even is, and their attention is short, so tiny, soft, frequent rewards suit them well. The same food budgeting rules apply, and with a puppy it matters even more to count treats, because a small dog and a generous hand can add up quickly.
Adult dogs often have more settled preferences, and you can lean on them. A dog who has been trained with food for years may respond beautifully to a toy or a game, and varying the reward keeps things interesting rather than letting a single treat go stale.
The principle does not change with age. Match the reward to the moment, keep it small, deliver it fast, and let the dog tell you whether it worked by how eagerly they offer the next repetition.
When food is not the answer at all
There are times when adding food is the wrong move. If your dog is frightened, guarding something, or mid-conflict with another dog, tossing treats does not fix the underlying feeling and can accidentally reward or mask it. Those situations call for management and professional help rather than a better treat pouch. Our guide to management and training explains where that boundary sits.
A sudden loss of appetite for food a dog normally loves can also be a sign something is off physically, and a vet is the right first port of call if that is new and unexplained. This guide is about everyday training rewards, not about diagnosing why a dog has gone off their food.
For everyday skills, though, food, play, and attention cover almost everything you need, and the whole art is choosing the right one for the job in front of you.
A simple decision you can run this week
Pick one behaviour you are currently working on and one new setting where it keeps falling apart. Before the session, decide your three tiers: what is ordinary, what is strong, and what is the jackpot.
Start the behaviour somewhere easy with the ordinary reward. Then move to the harder setting and switch to the strong reward the moment the difficulty climbs. If your dog still blanks, you have learned something useful: the setting is too hard for today, and the answer is more distance or a quieter spot, not more cheese.
This one experiment, run twice in a week, teaches you more about your own dog than any list of recommended treats, because the real answer is always the same: the reward that works is the one your dog genuinely wants in that moment, and your job is to notice what that is and have it ready.