How to Read a Kava COA in 60 Seconds (2026)

A Certificate of Analysis is the report an independent testing laboratory issues after it analyzes a specific lot of kava. The operative word is specific: a COA describes one batch, identified by a batch or lot number, on a particular date. It is not a brand-wide quality badge, and it is not the marketing copy on the bag. It is a snapshot of one production run, measured by someone who doesn't profit from the answer.

A complete kava COA carries five kinds of information, and a thorough one adds a couple more. The five that matter most: the testing lab and its accreditation; the batch/lot identity (and ideally the origin and processing date); the potency and profile — total kavalactones by HPLC, broken into the six major kavalactones, which yields the chemotype; the microbial panel; and the heavy metals panel. A good sheet also reports moisture (too-wet root invites mold) and, for kava specifically, a noble-status / flavokavain read that confirms you've got a traditionally consumed variety and not a tudei.

You don't read a COA top to bottom like a novel. You run five targeted checks, in priority order, and you can bail the moment one fails. Here's the sequence — the same one the 60-second check below formalizes.

1. The lab name and its accreditation. Who ran this test? A legitimate COA names the laboratory and shows an ISO/IEC 17025 accreditation — the international standard for testing-lab competence, typically issued through a body like A2LA or ANAB. This is check one because it qualifies everything beneath it: numbers from an unnamed or unaccredited lab aren't numbers you can trust, no matter how good they look. A brand testing its own product in-house, with no third party named, is the weakest possible version of this.

2. The batch number matches your product. A COA is only meaningful if it describes the bag in your hand. Find the batch or lot number on the certificate and confirm it matches the one on your package. A brand that posts a single COA and reuses it across every lot for years is showing you a photo of a different shipment. A batch number that doesn't match — or the absence of any batch number at all — is the most common and most disqualifying red flag there is.

3. Total kavalactones (by HPLC) and the chemotype. Now the potency. Look for total kavalactones, measured by HPLC (high-performance liquid chromatography — the accurate method), expressed as a percentage of the dry root. For noble kava root powder this commonly lands in the ~6–15% range. Then find the breakdown into the six major kavalactones; ranked most-to-least abundant, those become the chemotype code that tells you whether the kava skews heady or heavy. (Mechanics of the code live in Kavalactones, Explained.)

4. The microbial panel. Kava is a root; roots carry microbes. A clean COA shows tests for yeast and mold, E. coli, and Salmonella (often coliforms and total plate count too), each within an acceptable limit. The presence of this panel is itself a quality signal — and its absence is a warning. A potency number with no microbial testing beside it tells you how strong the kava is but nothing about whether it's safe to drink.

5. Heavy metals. Because kava draws up whatever is in its soil, a thorough COA screens for lead, arsenic, cadmium, and mercury, each reported against a limit. This is the last check because it's the least variable across reputable suppliers — but a missing heavy-metals panel on an otherwise complete sheet is still a gap worth noticing.

One number on a COA — or, more often, on the package quoting the COA — fools almost everyone, and it's the same trap we flag in Kavalactones, Explained: milligrams of root are not milligrams of kavalactones. The COA version of this trick is a percentage played in two different registers, and brands sometimes blur them on purpose.

Here's the confusion. A COA for a concentrated extract might honestly report "30% kavalactones." Impressive — until you realize that's 30% of the extract powder, and the finished product might be a capsule containing only a small amount of that extract cut with filler. "30%" describes the raw material, not your serving. The percentage that matters to you is kavalactones as a share of what you actually consume, or better, the milligrams of kavalactones per serving. A 30%-of-extract claim and a stated mg-per-serving figure are answering completely different questions, and the louder one is usually the less useful one.

Our rule is deliberately blunt, because the category rewards bluntness: no number, no value ranking. If a product cannot produce a batch-matched, third-party Certificate of Analysis with a real kavalactone figure, we do not place it on a ranked list. Not because we assume it's bad — we have no way to know — but precisely because we have no way to know. Ranking it would be guessing, and dressing a guess up as a recommendation is the opposite of what this site is for.

This cuts cleanly in practice. A brand that posts per-batch COAs is telling you it knows its material and will stand behind it — that's a point in its favor before you read a single result. A brand that's vague, that hides behind "proprietary blend," that quotes a percentage with no document attached, or that brags about milligrams of root while never stating kavalactones, has told you something too: that the real strength is unknown. In an unregulated aisle, demanding the number isn't pedantry. It's the buyer's main and best form of protection. The full standard is written up in How We Research.

Here's how the five checks land on an actual sheet. Every number below is illustrative — a realistic but invented noble-kava COA, built to teach the read, not lifted from any specific product. Imagine this printed on a one-page PDF you opened from a product listing.

Header. "Pacific Analytical Labs — ISO/IEC 17025 Accredited (A2LA Cert. #1234.01)." Check 1 passes: a named, accredited, independent lab. If this line read "Tested in-house" or named no lab at all, you'd stop here.

Sample identity. "Sample: Noble Kava Root Powder · Origin: Vanuatu · Lot #VK-2026-014 · Received: 2026-05-02 · Report date: 2026-05-09." Check 2: you find "Lot #VK-2026-014" on the bag in your hand — it matches. The dates are recent. Good. A months- or years-old report, or a lot number you can't match, fails here.

Potency (HPLC). "Total kavalactones: 9.8% (w/w) by HPLC." Check 3, part one: 9.8% sits comfortably inside the ~6–15% noble-root range — a solid, believable potency. Then the breakdown:

"Kavain 3.4% · Dihydrokavain 2.1% · Methysticin 1.2% · Dihydromethysticin 1.0% · Yangonin 1.4% · Desmethoxyyangonin 0.7%." Rank those most-to-least and read off the digits: kavain (4), dihydrokavain (2), yangonin (6)… giving a chemotype that opens 4-2-6. Check 3, part two: a code led by kavain (4) predicts a heady, uplifting, sociable kava rather than a heavy, sedating one. You now know the strength and the direction before the first shell.

Noble status. "Flavokavain B / noble verification: PASS (consistent with noble chemotype)." Confirms a traditionally consumed variety, not a tudei — the distinction Noble vs. Tudei is built around.

Microbial panel. "Yeast & mold: <100 CFU/g · E. coli: Negative · Salmonella: Negative (per 25 g) · Total plate count: within limit." Check 4 passes: the panel is present and clean. A sheet with no microbial section, however strong its potency number, fails this check by omission.

Heavy metals. "Lead: <0.5 ppm · Arsenic: <0.3 ppm · Cadmium: <0.3 ppm · Mercury: <0.1 ppm — all within limits." Check 5 passes.

Moisture. "Moisture: 8.2%." A sensible, dry-enough figure — high moisture is what lets mold take hold, so it's a quiet but real quality signal.