1 · Evidence grading
- A — multiple placebo-controlled trials in dogs, consistent results, objective outcomes.
- B — at least one well-designed placebo-controlled canine trial supporting the outcome.
- C — controlled evidence is negative, conflicting, or badly underpowered.
- D — credible signal of harm or clear failure across trials.
- I — insufficient: surveys, anecdotes, or extrapolation from other species only.
Grades use canine data only. Human results inform discussion but never a grade. Negative trials are reported with the same prominence as positive ones.
2 · Product verification (all five, or no listing)
- Published certificate of analysis, batch-matched to current inventory.
- Independent third-party laboratory — in-house testing doesn't count.
- Pet-specific formulation: no xylitol, no essential-oil additives toxic to pets, THC non-detect or <0.3%.
- Label accuracy: COA potency within ±10% of the label claim.
- Price-per-mg computed at list price and shown, so convenience formats can't hide their premium.
3 · The affiliate firewall
- Commissions never touch grades, verification outcomes, or rank order.
- Products that pay us nothing are listed when they pass; products that pay well are excluded when they fail.
- Affiliate links are marked
rel="sponsored"and disclosed above every comparison. - Brands cannot pay for placement, review, or removal of a negative finding. There is no price.
4 · Medical boundaries
Educational content only — never veterinary advice, never a diagnosis. Claim language stays inside the evidence ("reduced stress measures in one trial"), and FDA status is stated plainly: CBD is not FDA-approved for animals. Your veterinarian is the decision-maker; we're pre-reading.
5 · Maintenance
New canine trials get incorporated and grades revised — the footer's "last research review" date reflects the most recent pass. Prices re-checked on review; corrections logged on-page.