If you're already feeding your dog well and you're looking at raw food, you're not starting from zero. You've made one considered decision about nutrition already. This is the next one, and it should be simpler than the category currently makes it.
The difficulty is that raw is a hard category to read. Search "best raw dog food UK" and you'll find lists, but a list only helps if you know what the ranking was based on. The brands in the category are not all doing the same thing.
So this isn't a ranking. It's four questions worth asking about any raw food, plus one question to ask about every answer you get. Which of the four matters most depends on your household. For some people freezer space decides it before anything else does. Use it on any brand you're considering, including us.
The four questions are: is it right for your dog, can you feed the right amount, how is the safety risk handled, and will it work in your house. The fifth question runs underneath all of them. Anyone can tell you the answer. What can they show you?
1. Is it nutritionally appropriate for your dog?
Raw food is not complete by default. Muscle meat alone doesn't supply everything a dog needs, and a food is complete when its composition meets the dog's requirements, not because it contains any particular list of ingredients.
Two things are worth establishing, and they're different.
Is it complete, and for which dog? "Complete" is a defined term in UK pet food regulation. It means the food supplies a dog's full daily nutritional needs. A complete food that doesn't name a life stage should be formulated for all of them. If it's made for a specific life stage, the label should say so.
That matters because a growing puppy is building bone and tissue rather than maintaining it, and a pregnant or nursing dog is feeding more than herself. Both need more of several nutrients than an adult at rest. In Europe those levels are published by FEDIAF. A brand with a range should be able to point you at the right food in it. A brand with one recipe should be able to tell you which life stages it was formulated for.
Where does the nutrition come from? Micronutrients like iodine, manganese, zinc and vitamin D have to get into the food somehow. Some brands use whole-food ingredients chosen for what they contribute. Some use a vitamin and mineral premix. The two approaches can also be combined. All of these are legitimate, and a food built on a premix is not a worse food.
You don't need to audit the ingredient list to check this. You need the brand to be able to explain its own recipe. A brand that has done the formulation work can tell you in a sentence how the micronutrients are covered. One that can't is telling you something.
Ask: "Is this complete for my dog's life stage, and how are the vitamins and minerals covered?"
They should be able to show you: the basis for the claim. An analysis of the finished food by a competent laboratory is strong evidence, and so is the basis of the formulation together with appropriate ongoing controls. A declaration on the label with nothing behind it is neither.
2. Can you feed the right amount, consistently?
Vets estimate around half of dogs in the UK are overweight. Around four per cent of owners think their own dog is. Weight has many causes and portioning is only one part of managing it, but it is harder than it looks.
Getting it right needs two numbers.
Your dog's energy requirement. One method is a percentage of body weight. It's easy to explain and it takes no account of the dog. A neutered older Labrador with a calm temperament and a much more active intact younger one can weigh the same and need noticeably different amounts.
The alternative starts from Resting Energy Requirement and applies adjustment factors for age, activity, neuter status and body condition. This gives a more individualised starting estimate. It's worth being clear about what it doesn't give you. An equation predicts the average dog with those characteristics, and your dog may sit either side of that. The published guidelines describe their own energy figures as starting points for exactly this reason.
The food's energy density. Knowing your dog's calorie requirement tells you nothing until you know how many calories are in the food in front of you. Raw recipes vary in energy density, and a portion in grams is only meaningful against the recipe it was calculated for. If a brand rotates recipes, this matters more rather than less. One gram figure can't deliver the same calories across recipes of different energy densities unless the brand has accounted for the difference somehow, so it's fair to ask how.
Then you watch the dog. Any calculation is a starting point. The portion that's right is the one that keeps your dog's weight moving the way it should over the following weeks, which for an adult means holding steady and for a puppy means growing at a sensible rate. So the useful thing a brand can give you is a number to start from and a method for adjusting it.
There's a practical note here too. Measuring by eye or by scoop is unreliable. When 100 owners were asked to measure out dry food using cups and scoops, the results ranged from forty-eight per cent under the target to a hundred and fifty-two per cent over, with the worst accuracy at the smallest volumes. Kitchen scales avoid that volume-measurement problem, and a pack sized to a day's feed avoids it a different way. Either is fine. Guessing isn't.
Ask: "How did you arrive at this number, and what's the calorie content of the recipe you're sending me?"
They should be able to show you: the energy density of that specific recipe, and a feeding method that asks for more than your dog's weight.
Our feeding calculator works this way. The calorie target it gives you is your dog's, so it's useful whichever brand you choose. The gram figure isn't: that's the calorie target converted using our recipe's energy density, and converting it for another food needs that food's figure instead.
3. How is the safety risk handled?
This is the question that's specific to raw, and it deserves a straight answer rather than reassurance.
Raw meat can carry pathogens. In a 2026 survey, the Food Standards Agency tested 380 raw pet food samples from 50 brands. Around 29 per cent exceeded the statutory microbiological criteria that apply to them, and Salmonella was detected in around a fifth. The survey covered 50 brands, not one manufacturer or an isolated recall.
Raw pet food isn't heat-treated to eliminate bacteria during production. That's the honest starting point: the risk is managed rather than removed, and a brand that tells you otherwise is telling you something that isn't true of the format.
What can be managed is a great deal. Contamination can arise from incoming material, the production environment, equipment or handling. A brand's past results tell you something real about how well its process performs over time, but they don't establish the status of the next production run. That's why the controls worth asking about are the ones that operate on every run rather than on the recipe.
So it's worth understanding what a brand actually does, in this order.
Before production. Where the meat comes from and what's required of suppliers. Approved premises, incoming checks, cold chain.
During production. Hygiene controls and process discipline. In Great Britain, a plant making pet food from animal by-products needs APHA approval, which includes an inspection of the site. That's the regulatory floor, not evidence that one brand's safety system beats another's, and not every brand on a UK shelf is made here.
Before dispatch. Testing. Three details separate systems that look similar on a website. Is every batch tested, or a sample of batches periodically? Testing every batch means every run gets sampled; periodic testing leaves some runs unsampled. Is the batch then held until the result comes back, or dispatched while the laboratory works? Holding is what turns the test into a gate before release rather than a check after it. And who does the testing, with what accreditation for the methods used?
One honest limit, because it matters for reading anyone's claims including ours. Testing takes samples from a batch. A pass means the samples met the release criteria, which reduces the uncertainty considerably. It is not a statement about every gram, and any brand presenting it as one is overstating what a lab result can do.
Ask: "Is every batch tested before it's dispatched, is it held until the results are back, and who does the testing?"
They should be able to show you: how the controls work and what testing is done, in a form that doesn't rest on their own word alone. Independence adds weight here, and the strongest form is a result for the batch you were actually sent, by name and date. A brand without it is not automatically failing; a brand with it has reduced how much of the safety claim rests on its own word.
4. Will it work in your house?
The best food in the category is no use if feeding it is a project. This is the part that can go unexamined until the first delivery arrives.
Start with freezer space. A month of food for a large dog is a real volume, and it has to go somewhere that isn't already full. Worth working out before you order rather than after the delivery arrives.
Then the daily loop. Frozen food needs defrosting, and how far ahead depends on the format and the portion size. It's a habit rather than a task, but it's a habit that has to survive a Tuesday when you forgot. Ask how long it needs and what the fallback is when you don't manage it.
Then delivery. How often, in what, and does someone need to be in. The question worth asking is how long the packaging holds temperature and what the brand tells you to do if a box has been sitting out longer than that.
None of this is glamorous and none of it appears in a comparison table, which is exactly why it's worth asking about. A food you can feed correctly on an ordinary weekday beats a marginally better food you end up rushing.
Ask: "How much freezer space does a month take, and what does the daily routine actually look like?"
They should be able to show you: the pack dimensions and a straight description of the routine, not a claim that it's effortless.
The question underneath all four
A brand can answer all four questions well without any of the answers being checkable. The useful distinction isn't the answer. It's what sits behind it.
There's a ladder here, and it's worth knowing which rung you're being offered.
The principle is that the evidence should fit the claim, and what counts as good evidence is different in each case.
For nutrition it's the basis of the formulation, an analysis of the finished food using appropriate analytical methods, and controls that continue as ingredients vary. Independence from the brand adds weight to any of it, but competence with the method is the part that has to be there. Microbiological testing is different: for raw pet food an accredited laboratory is part of the regulatory requirement, not an optional upgrade. For portioning it's the published energy content of the recipe you're being sent, and a method you can follow. For safety it's the controls plus testing, and the strongest version is a result for your specific batch. For practical fit it's the pack dimensions and the delivery conditions, which need no third party at all.
That last one is worth noticing. Independent evidence isn't a universal good. It's the right answer where a brand's own word is the weak link, and unnecessary where it isn't.
There is a structural reason to ask. Under the ordinary feed-claims rules, a claim like this isn't authorised by anyone before sale. The business makes it, and the scientific substantiation has to exist when the food goes on the market and be produced if the competent authority asks for it. So the substantiation is required to exist. It just isn't required to be shown to you. That gap is where the question lives, and it's a feature of how the market is regulated rather than a sign of a bad brand.
But it does mean the question is worth asking every time, of every claim, including ours.
Putting it together
The trade-offs are real. A brand can have done careful formulation work and test infrequently. It can test every batch and portion by percentage of body weight. It can do both and arrive in a format that fills your freezer.
Working through the four tells you what you'd be accepting. Asking what sits behind each answer tells you how much of it you're taking on trust.
For how costs compare across formats, see How Much Does Raw Dog Food Cost?. For the practical side at home, see How to Store and Handle Raw Dog Food Safely. For a longer look at completeness, see Is Raw Dog Food Nutritionally Complete?