The line that actually tells you whether a cat food is complete and balanced is the AAFCO nutritional adequacy statement, printed in small type on the back panel, and the guaranteed analysis above it means nothing until you subtract the moisture and convert to a dry matter basis. Almost every comparison people make in the aisle happens on as-fed percentages, which count water as though it were food. A canned food at 10% protein and a kibble at 32% protein are not the same kind of number, and once the water comes out, the ranking very often flips.
I keep the calculator app in the third slot on my phone's home screen for exactly this, which sounds strange until you've stood in a narrow pet store aisle with a can in each hand and your thumb resting on the moisture line. Fluorescent light overhead, bag of litter wedged under one arm. The front of the can says nothing useful. The back of the can, bottom third, says everything, as long as you know which number is the one doing the misleading.
Read the first can. Crude protein, minimum 10%. Crude fat, minimum 6%. Crude fiber, maximum 1.5%. Moisture, maximum 78%. That last one is the whole trick, and it is the only line on the label that never appears in the marketing.
Do the arithmetic with your thumb. If 78% of the can is water, then 22% of it is actual food. The protein is 10% of the entire contents, water included, so divide 10 by 22. You get about 45. Forty-five percent protein on a dry matter basis.
Now the bag. Protein minimum 32%, moisture maximum 10%. Ninety percent dry matter. Thirty-two divided by ninety is about 35.5.
The can that advertised 10% is the denser food. The bag that advertised 32% is not. Both labels are legal and neither one is lying. They're answering a question you didn't ask, and the aisle is built on the assumption that you'll line them up side by side and reach the wrong conclusion.
The guaranteed analysis is a floor and a ceiling, not a measurement
Nothing on that panel is an exact amount. Protein and fat are minimums. Fiber and moisture are maximums. A food guaranteed at 30% protein contains at least 30% and could contain considerably more, because manufacturers build in margin so that no batch ever tests under the legal floor and gets them in trouble. You are reading the edges of a range, not the middle of it.
That's why the panel is so easy to game. Fiber is cheap; fat and protein are not. So fiber gets a maximum and fat gets a minimum, and the numbers are structured to keep the manufacturer compliant rather than to inform you.
Two more things about that panel that took me a while to understand. Crude protein is not protein. It's a nitrogen measurement multiplied by a standard factor, which means it can't distinguish chicken from peas from corn gluten, and it tells you nothing about how much of that protein the cat can actually digest. And ash, when it's listed at all, is part of the dry matter too. It isn't protein, fat, fiber, or water, but it occupies space inside that 22%, so it belongs in your accounting.
The practical move here is the one almost nobody makes: email or call the manufacturer and ask for the typical analysis, not the guaranteed one, plus the calorie content. A lot of them will send you a PDF. The ones who won't have told you something as well.
What the AAFCO statement is, and what it is not
This is the sentence near the bottom, often next to the phone number, and it comes in a few standard forms. One says the food is formulated to meet the AAFCO nutrient profiles for a named life stage. Another says animal feeding tests using AAFCO procedures substantiate that the product provides complete and balanced nutrition for a named life stage. A third, and this is the one to slow down for, says the food is intended for intermittent or supplemental feeding only, which means it is not complete and balanced at all.
Both of the first two are legal. They are not equal.
The formulation route proves that the recipe, calculated on paper, hits a nutrient table for a given life stage. It's cheaper and it's the only realistic option for a company with dozens of products, and plenty of good foods use it. The feeding trial route puts the finished food in front of real cats for roughly six months, on a minimum of eight of them, and shows that they maintained condition on it. That's the gold standard, and it's the single strongest signal you can find on a label, because it's the only claim that tested the food as eaten rather than as calculated.
What the statement does not tell you is anything about ingredient quality, digestibility, or whether your cat will touch it. "Complete and balanced" is a nutrient adequacy claim, no more. And note the life stage. "All life stages" means the food meets the stricter growth standard, which is not the same as saying it's the right choice for a sedentary adult, just that it clears a higher bar.
The ingredient list is a ranking, not a recipe
Ingredients are listed by weight before processing. That single fact explains most of what looks strange on a label.
Fresh chicken is roughly seventy percent water. It goes into the mixer heavy and it comes out of the oven light. Chicken meal is already rendered and concentrated, so a food listing chicken first and chicken meal fifth may well contain more chicken meal than chicken in the finished product. The list isn't wrong. It's just measured at a moment that flatters the marketing.
The word "meal" has a specific definition and it isn't a slur. Neither is "by-product," which refers to clean parts other than the muscle meat, and can include organ meats that are genuinely nutrient-dense. "Natural" has a definition too, but it's about sourcing and synthetic processing, not quality. "Holistic" has no regulatory definition at all, which is why you see it on everything. "Human-grade" speaks to the supply chain and the facility, not to whether the food meets a nutrient profile.
Then there's splitting. A list that reads chicken, brown rice, ground corn, corn gluten meal, wheat gluten, chicken fat puts chicken at the top, but the two plant proteins counted together may outweigh it. I'm not going to tell you plant proteins are bad. Some cats do fine on them and the amino acid profile matters far more than the source, taurine being the obvious one. The point is narrower than that: the order of the list is being used to sell you something, and the break-apart trick is how.
The number that beats both of them
Dry matter basis gets you out of the water trap. It doesn't get you out of the calorie trap.
Cats eat to a calorie need, not a volume need. Two foods can sit at the same dry matter protein percentage and deliver different amounts of protein per bite, because one is far more calorie-dense than the other. The comparison that survives that is grams of protein per 100 kilocalories.
It's easy to run and the numbers are already on the label. Take the protein percentage, multiply by the weight of the can to get grams. Take the kcal per can, usually printed near the adequacy statement as kcal ME/kg or kcal/cup. Divide grams by calories and multiply by 100. A three-ounce can at 10% protein gives you about 8.5 grams; if the can is 80 calories, that's roughly 10.6 grams of protein per 100 calories. Do that for both foods and the picture stops being ambiguous.
The order I actually do this in
Flip to the back first and find the AAFCO statement. Life stage, feeding trial or formulated, and whether it says intermittent feeding. If it says intermittent, the rest doesn't matter for a daily diet.
Then the moisture line, then the calculator. Then, and only then, the ingredient list, and only the first five entries. Then the calories per can or per cup. The front of the package gets none of my attention, because everything on it is a word someone chose.
I do the whole thing in about ninety seconds now. It took me a couple of years of keeping a spreadsheet of foods my fosters would actually eat to get that fast, and the spreadsheet is the part I'd recommend to anyone, because the label can't tell you the only thing that ultimately matters, which is whether the cat cleans the bowl.
What this method will not tell you
It won't tell you digestibility or amino acid availability. It won't tell you batch-to-batch consistency. It won't tell you palatability, and a food that's flawless on paper and refused is worth nothing at all. It won't tell you whether the AAFCO nutrient profiles themselves are optimal, because they're a floor for adequacy, not a target for ideal, and nutritionists genuinely disagree about where above that floor you should aim.
And it isn't a medical tool. If a cat has a diagnosed condition, the label is the wrong instrument entirely.
I put the bag back on the shelf. Not because it was a bad food. I have no idea whether it's a bad food, and neither does the bag. I put it back because it told me 32% and meant 35 on a dry basis and something less than that per hundred calories, and I had two cans at home that beat it on both counts and carried a feeding trial statement. The one I bought had 78% moisture and a trial behind it, and there's a fair chance the cat will sniff it and walk away. That's the one thing no panel has ever been able to print.
