← Blog·Nutrition Science

Dog Food Ingredients to Avoid, and the Myths to Ignore

September 6, 2026 · 13 min read

Close overhead view of a person's hands holding a dog food package and studying the fine-print ingredient panel under warm kitchen light, with a pair of reading glasses and a notepad on the counter.

Short answer: the evidence-backed avoid list is much shorter than the internet suggests. Xylitol, spoiled or mouldy food, and whatever your individual dog has a diagnosed reaction to. That is close to all of it. Corn, by-product meal, and meat meal, the three names that appear on nearly every viral blacklist, are safe and often nutritionally useful.

This will read as contrarian. It is the mainstream position among veterinary nutritionists, and it is worth understanding why the popular version drifted so far from it.


The list that holds up

Xylitol

Genuinely dangerous, and the one item here with no nuance attached. Xylitol triggers a rapid insulin release in dogs, producing severe hypoglycaemia within 30 minutes, and higher doses cause liver failure.

It is not an ingredient in commercial dog food. The risk is in human products that end up shared: sugar-free peanut butter, gum, baked goods, some supplements. Read the label on anything from your own cupboard before it becomes a treat.

Spoiled, mouldy, or long-expired food

Mycotoxins from mould growth on grains and on food left in a damp bin are a documented cause of illness and have driven real recalls. So is rancid fat, which develops in any bag left open too long and destroys vitamin content as it goes.

This is a storage problem more than a shopping problem. Keep food in its original bag inside a sealed container, buy a size your dog finishes in about six weeks, and throw out anything that smells sharp or looks off.

Whatever your dog is actually reactive to

Food allergies in dogs are real but less common than the market implies, and they are overwhelmingly to proteins the dog has eaten a lot of: beef, dairy, chicken, wheat. Not to exotic villains.

The point is that this list is specific to your dog and requires a diagnosis. Guessing at it by eliminating ingredients you read about is how owners spend a year on rotation diets without ever running the elimination trial that would answer the question. Our dog food allergies guide covers how the work-up is actually done.


The myths, and where they came from

"Corn is a cheap filler"

Corn supplies digestible carbohydrate, linoleic acid, protein, and antioxidants, and cooked corn in dog food is well digested. It is not padding in any technical sense; a filler would be an ingredient with no nutritional contribution, and corn does not qualify.

Corn allergy exists. It is uncommon compared with beef, dairy, and chicken.

The filler story was manufactured. It came from brands selling grain-free alternatives that needed a villain, and it stuck because it sounds like common sense.

"By-products are floor sweepings"

AAFCO's definition of meat by-products covers clean, non-rendered parts other than muscle meat: liver, spleen, kidney, lung, blood, bone. The same definition explicitly excludes hair, horns, teeth, and hooves.

Those organ tissues are nutrient-dense, frequently richer in vitamins and trace minerals than muscle meat, and they are what a wild canid eats first when it opens a carcass. A food using them is making a defensible nutritional choice.

The revulsion is cultural. We do not eat spleen, so we assume feeding it is an insult.

"Fresh chicken listed first means more meat"

This one inverts. Ingredient lists are ordered by weight before cooking, and fresh chicken is about 70% water. Once the kibble is extruded, most of that weight is gone.

Chicken meal is chicken with the water already removed, which makes it roughly four times more protein-dense by weight. So a food listing chicken meal third can easily contribute more actual protein to the finished product than one listing fresh chicken first.

A label leading with fresh meat often reflects marketing awareness rather than a better formula.

Watch for peas, pea protein, pea flour, and pea fibre listed separately. Split across four entries, each lands below the meat. Combined, they might outweigh it. The list is being composed to read a particular way.

"BHA and BHT cause cancer"

Both are approved antioxidant preservatives that stop fat going rancid, which is itself a genuine health risk. The concern traces to rodent studies using doses far above what pet food permits, and to forestomach tumours in a species with an organ humans do not have.

If you would rather avoid them, mixed tocopherols are a vitamin E based alternative, with a shorter shelf life as the tradeoff. Treat this as a preference. It should not outrank who made the food.

"Grain-free is safer"

The opposite concern currently applies. The FDA has an open inquiry into diet-associated dilated cardiomyopathy in dogs eating grain-free formulas heavy in peas, lentils, and other pulses. Causation is not established, and the picture is still unsettled.

Grain-free is a descriptive choice rather than a health upgrade. Choose it for a diagnosed medical reason, not by default. The grain-free and heart disease guide has the current state of the evidence.


Flags that are worth something

A short list of things that do tell you something, for narrower reasons than usually given.

Generic protein terms. "Meat and bone meal" or "animal fat" without a species is not dangerous, but it signals the manufacturer has left room to change suppliers between batches. That variability matters if you are managing a suspected sensitivity and need consistency.

No AAFCO statement, or the weaker version of it. A food that says "formulated to meet" was checked on a spreadsheet. One that says "animal feeding tests" was fed to actual dogs. The second is stronger evidence. Our label guide shows exactly where to find this.

No published kcal/kg. A manufacturer that will not tell you the calorie density is making it impossible to portion correctly or compare prices.

Nutritional claims with no formulator behind them. If a company cannot say who designed the recipe and what their credentials are, the ingredient list is decoration.


What actually predicts food quality

Not the ingredient panel. The company.

The WSAVA criteria ask whether a manufacturer employs a full-time board-certified veterinary nutritionist, uses a credentialled formulator, runs AAFCO feeding trials, owns its plants, and operates a documented quality-control programme.

Those five questions separate foods far more reliably than any ingredient list, and they explain a pattern that confuses people: some of the most expensive boutique foods, with the most appealing ingredient panels, come from companies that answer no to all five. Meanwhile several plain-looking mid-priced foods answer yes to all five.

Ingredient-list scoring is exactly what star-rating sites do, and it is why those ratings mislead. A list can be composed to score well. Nutritional competence cannot.


The bottom line

Avoid xylitol, spoiled food, and your own dog's diagnosed triggers. Stop avoiding corn, by-products, and meat meal, which are safe and useful. Notice generic protein sources and missing AAFCO feeding-trial language, then spend your attention on the manufacturer rather than the panel.

The best-formulated food in the country would still fail an ingredient-list purity test somewhere. That tells you the test is measuring the wrong thing.

Get a shortlist built on manufacturer quality, not ingredient panels →

FAQ

Quick answers sourced from veterinary literature

These mirror the medically reviewed IntelliBowl notes on this slug and exist to help crawlers summarize quotable excerpts.

Which dog food ingredients should I actually avoid?

Xylitol, which causes severe hypoglycaemia and liver injury in dogs and appears in some human foods rather than commercial dog food. Anything visibly mouldy or past date, since mycotoxins are a real and documented hazard. And any protein your dog has a veterinary-diagnosed adverse reaction to. Nearly every other item on a typical internet avoid list has no evidence behind it.

Is corn bad for dogs?

No. Corn is not a filler in any meaningful sense; it supplies digestible carbohydrate, linoleic acid, and protein, and cooked corn in dog food is well digested. Corn allergy exists but is uncommon compared with reactions to beef, dairy, and chicken. The idea that corn is a cheap padding agent comes from marketing by brands selling grain-free alternatives, not from nutritional science.

Are by-products bad in dog food?

No, and the term is widely misunderstood. AAFCO defines poultry by-products as clean parts other than meat: organs such as liver, spleen, and kidney. Those tissues are nutrient-dense, frequently richer in vitamins and minerals than muscle meat, and are exactly what a wild canid eats first. The regulated definition excludes hair, horn, teeth, and hooves. The revulsion is cultural rather than nutritional.

Is chicken meal worse than fresh chicken on a label?

Usually it is the more informative ingredient. Chicken meal is chicken with the water already removed, so it is roughly 300% more protein-dense by weight than fresh chicken. Because the ingredient list is ordered by pre-cooking weight, fresh chicken listed first can contribute less actual protein to the finished kibble than a meal listed third. A label leading with fresh meat often reflects marketing awareness rather than a better formula.

Are BHA and BHT in dog food dangerous?

At the concentrations permitted in pet food, the evidence for harm is weak. Both are approved antioxidant preservatives, used at levels far below those in the rodent studies that generated the concern, and they prevent fat rancidity, which is itself a genuine health risk. Mixed tocopherols, a vitamin E preservative, are a reasonable alternative if you prefer one, with the tradeoff of a shorter shelf life. Neither choice should outrank manufacturer quality in your decision.

Should I avoid foods with unnamed meat sources?

This one is a fair signal, though for a narrower reason than usually given. Generic terms like meat and bone meal or animal fat are not dangerous, but they tell you the manufacturer has left itself room to change suppliers between batches. That variability matters if you are managing a suspected food sensitivity and need consistency. For a healthy dog eating a food from a company that runs proper quality control, it is a minor consideration.