The State of Pet Food — What 2,354 Verified Labels Show

Published July 16, 2026 · Corpus data as of July 11, 2026 · By Erin Rose

We pulled guaranteed analysis, calorie content, ingredients, and price off 2,354 official dog and cat food labels (1,112 dog, 1,242 cat, across 50 brands) and ran the numbers brands don't publish. The headline findings: wet dog food costs 5.7× more per calorie than dry; dry cat food runs roughly 30% carbohydrate on a dry-matter basis versus about 5% for wet, despite cats being obligate carnivores; 20.3% of recipes carry two or more legumes in their first ten ingredients, the pattern under the FDA's open grain-free/DCM inquiry; label composition and bag size together explain only 43% of what dry dog food actually costs; and the grain-free premium isn't one number — it runs positive in some segments and negative in others. Every figure below is computed directly from the corpus, not estimated.

What a calorie costs

Cost per bag or cost per pound doesn't account for calorie density, so it isn't a fair way to compare foods. Cost per 1,000 kcal is: it's the only unit that lets you compare a dense kibble to a watery can on a like-for-like, "what does it cost to actually feed this animal" basis. See our methodology for the full formula.

Macronutrient reality

Guaranteed analysis is published "as fed," which bakes in each food's moisture content — a wet food's numbers aren't directly comparable to a dry food's until you convert both to a dry-matter basis. Once you do, the obligate-carnivore angle for cats gets sharper: dry cat food's dry-matter carbohydrate median (30.0%) is roughly 7× the wet-food median (4.6%) — dry kibble needs starch to hold its shape; a cat's biology doesn't need the carbohydrate that comes with it.

The grain-free premium

"Grain-free costs more" isn't a single fact — it depends on species and form. In this corpus, grain-free carries a premium in dry dog food and wet cat food, but grain-free is actually cheaper at the median in dry cat food and wet dog food. Grain choice isn't a reliable price signal on its own; check the actual cost per 1,000 kcal for the specific food.

The legume question (DCM context)

The FDA has an open inquiry into a possible association between certain grain-free diets and canine dilated cardiomyopathy (DCM); the inquiry's own reporting has focused attention on recipes heavy in peas, lentils, and other legumes rather than on the absence of grain itself. No causal mechanism has been established, and AAFCO-complete grain-free food is not inherently unsafe — this is an open scientific question, not a settled one. What we can measure directly is legume load: how many of the first 10 ingredients are legumes.

N = 2,186 recipes with a disclosed ingredient list.

What the label explains — and what you're paying for

We ran a regression of price per 1,000 kcal against what's actually on the label: protein, fat, and fiber (dry-matter basis), plus grain-free, novel-protein, and single-animal-protein flags, plus bag size for dry segments. The result (R²) tells you what share of the price difference between two foods is explained by their composition on paper. What's left over — the residual — is brand: sourcing, safety testing, manufacturing, research, marketing, everything that isn't printed as a guaranteed-analysis number.

Dog · Dry (n=410)

Label composition and bag size explains 43% of the price variation in this segment (R² = 0.432) — the rest is brand, sourcing, and things not printed on the label.

Dog · Wet (n=178)

Label composition explains 27% of the price variation in this segment (R² = 0.273) — the rest is brand, sourcing, and things not printed on the label.

Cat · Wet (n=382)

Label composition explains 24% of the price variation in this segment (R² = 0.237) — the rest is brand, sourcing, and things not printed on the label.

Method: OLS of log(cost/1,000kcal) on protein/fat/fiber (DM), grain-free, novel-protein and LID flags, plus log bag size for dry segments; brand premium = mean residual per brand (>=8 priced SKUs).

Value leaders

For each segment above, these are the foods whose actual price falls furthest below what their own label composition predicts — the biggest gaps between what the label says and what the food costs. This is not a quality ranking; it says nothing about taste, sourcing, or safety testing. Linked foods route to their full Verified Label page.

Dog · Dry

Dog · Wet

Cat · Wet

Transparency leaderboard

We score every SKU on how much of the intel we track (guaranteed analysis fields, calorie data, ingredient list, AAFCO statement, sourcing detail) the brand actually discloses on the label. Averaged per brand, that produces a transparency percentage — nothing to do with food quality, purely how much a brand tells you.

At the other end: Dr. Elsey's (33.3%), Rachael Ray Nutrish (33.3%), Hill's Science Diet (20.0%), Solid Gold (13.3%), and Sheba (0.0%) disclose the least of what we track, across 319 labels combined.

Methods & sources

• Guaranteed analysis, calorie content, ingredients, and AAFCO statements are pulled from official pet-food labels or label-deck PDFs, per our methodology.
• Cost per 1,000 kcal figures come from Amazon retail prices matched to a specific SKU; each priced product page states its own price-verification date, not one uniform snapshot for this report.
• All statistics on this page are descriptive of the current corpus (2,354 SKUs, 50 brands) as it stands today — not a claim about the full U.S. pet-food market. Denominators are stated per table; not every SKU discloses every field (92.9% have ingredients, 78.1% have calorie data, 54.6% are fully priced).
• The full corpus is available machine-readable at /dataset.json (CC-BY-4.0).