Formaldehyde Emission Limits for Furniture: 6 Costly Myths

Formaldehyde emission limits for furniture changed in the European Union on 6 August 2026, and the E1 certificate in your compliance folder no longer proves what your buyer thinks it proves. Since that date, REACH Annex XVII entry 77 caps formaldehyde released from furniture and wood-based articles at 0.062 mg/m³ — measured on the finished article, in a chamber, not on the raw panel your supplier tested.

That number is exactly half the E1 ceiling of 0.124 mg/m³, and it is measured under the same chamber conditions EN 717-1 uses: 23 °C, 45 % relative humidity, a loading factor of 1 m²/m³ and one air change per hour. Same test box, half the allowance, and the article on the hook instead of the board.

Most factories handle this by forwarding the panel mill's certificate and hoping. Then a container sits, a compliance team asks for a document nobody prepared, and a purchase order slips a month. Six beliefs cause most of it. Each sounds reasonable. Each is wrong in a specific, checkable way.

Myth 1: "We are E1, so Europe is covered"

Why people believe it. E1 has been the European baseline for two decades. EN 13986 defines two emission classes for wood-based panels, E1 and E2, and E1 became the effective market requirement across the EU from 2006. Every panel invoice says E1, so E1 feels like the finish line.

The truth. E1 is a panel class. Entry 77 regulates the article — the finished chair, the assembled wardrobe, the flooring plank as sold. Three panels that each pass E1 comfortably can still put a bookcase over 0.062 mg/m³ once they are glued, edge-banded and boxed together, because emission scales with how much board surface is in the room.

The limits that now apply:

Entry 77 carves out articles where formaldehyde is naturally present in the material, articles for exclusively outdoor use, articles exclusively for industrial or professional use where the public is not exposed, second-hand articles, and goods already covered by other regimes (food-contact articles, medical devices, personal protective equipment, biocidal products).

Germany got there first and is the reason "E1 is fine" started failing before 2026. Since 1 January 2020 the German Chemicals Prohibition Ordinance has used EN 16516 as its reference test at 0.1 ppm, and an EN 717-1 result must be multiplied by two before it is compared — which halves the effective EN 717-1 ceiling to 0.05 ppm. Suppliers who cleared German retail in 2019 on an E1 report found out in 2020 that the same report failed. The EU-wide rule has now done the same thing to everyone else, and it stacks on top of the wood-sourcing paperwork covered in EUDR furniture requirements.

Myth 2: "Our CARB P2 certificate covers the United States"

It did — until 22 March 2019. Before that date, a panel or finished good certified to CARB ATCM Phase 2 could be labelled either way, and the CARB label satisfied the federal requirement.

Since 22 March 2019, every regulated composite wood panel and every finished good manufactured in or imported into the United States must be labelled TSCA Title VI compliant. The relationship is now one-directional: California accepts goods labelled TSCA Title VI compliant, and CARB treats them as meeting its requirements including the ones that are stricter than the federal rule. A CARB-only label going the other way does not work.

There is one more label most exporters have never written: panels imported or moved purely for testing must be marked "for TSCA Title VI testing only, not for sale in the U.S."

Myth 3: "TSCA Title VI and CARB Phase 2 are the same rule"

The numbers are identical. The obligations are not, and California's own comparison table says so line by line.

The shared emission standards:

Where the two rules diverge:

Identical limits, different paperwork — and it is the paperwork that stops containers.

Myth 4: "A small composite-wood part does not need a label"

This one is half true, which is why it is expensive. Under the federal rule, a finished good does not need a label if its composite wood content is no more than 144 square inches measured on the largest face — a small picture frame, for instance. The wood itself still has to comply. Under CARB, there is no such exemption: if a finished good contains regulated composite wood, it gets a label.

So the safe rule for anything that might reach California is: label it anyway.

The two label types carry different fields, and mixing them up is an easy way to get a document review kicked back:

Panel label (on the panel or the bundle)

• Name of the panel producer
• Product lot number
• Compliance level — TSCA Title VI, or NAF / ULEF
• The EPA TSCA Title VI third-party certifier (TPC) number

Finished-goods label (on the good or the box)

• Fabricator name
• Production date, month and year
• Compliance level — and, if true for every composite wood product inside, a note that it was made with NAF or ULEF material

Note what is not on the finished-goods label: no lot number, no TPC number. Adding fields is not a safe hedge; it invites questions you then have to answer. The same discipline applies to any other mark you print — see CE marking size requirements for how precise the geometry of a compliance mark has to be.

Myth 5: "E0 and ENF are international grades"

E0 and ENF are real, defined, and Chinese. GB/T 39600-2021, in force since 1 October 2021, grades formaldehyde emission from wood-based panels and finishing products in three tiers — E1 at 0.124 mg/m³ or less, E0 at 0.050 mg/m³ or less, and ENF at 0.025 mg/m³ or less — and it retired the old E2 grade entirely.

Outside China, "E0" is a market convention rather than a standard: EN 13986 knows only E1 and E2, and the E0-style tiers that circulate in Europe trace back to private buyer specifications — IKEA's half-of-E1 internal class is the one the literature usually names. So an ENF claim is precise in Shanghai and needs translating in Hamburg. Write it as a number and a test method, not as a letter grade.

Myth 6: "F☆☆☆☆ is the Japanese version of CARB P2"

Different test, different unit, no legal bridge. Japan's F-star system is measured by the JIS A 1460 desiccator method in mg/L of solution, not ppm in a chamber. F☆☆ sits around 1.5 mg/L, roughly where E1 sits; F☆☆☆☆ is 0.3 mg/L or less.

Published correlations put F☆☆☆☆ near 0.035 ppm on the American large-chamber method — below the US hardwood-plywood limit of 0.05 ppm. That is useful for planning which mill to buy from. It is not a substitute for certification: no authority accepts one scheme's certificate in place of another's, and correlation tables are laboratory guidance, not legal equivalence.

Formaldehyde emission limits for furniture, side by side

Read that table once and the pattern is obvious: everyone is converging on roughly half of E1, by four different measurement systems, on four different clocks.

The fix: a spec sheet that answers the question before it is asked

The panel mill sends a certificate. The certificate goes into a folder. The buyer's compliance officer never sees the folder — they see your product page, your PDF catalogue and your quotation, and none of those say a word about emissions. Formaldehyde emission limits for furniture are lost at that hand-off far more often than they are lost in the lab: it is a documentation problem, not a chemistry problem.

Before the next purchase order goes out:

• Confirm which rule the destination applies — EU entry 77, US TSCA Title VI, California, China, Japan — and write it on the quotation
• Get the panel certificate and the test method and date, not just the class letter
• For the EU, ask whether the finished article has been chamber-tested, not only the board
• Check the label fields against the right list — panel labels and finished-goods labels are not interchangeable
• Label small composite-wood parts anyway if California is in the shipping plan
• Keep bills of lading, invoices and compliance statements for three years, not two
• If you make veneered tabletops or cabinet doors for the US, check whether you are now treated as a panel producer
• Put the emission class, the test method and the board thickness on the same drawing the buyer already looks at — see how a furniture spec diagram carries specification fields alongside measured dimensions

FAQ

Does an E1 certificate still mean anything after 6 August 2026?

Yes, but less than it used to. E1 remains the EN 13986 class for panels and it is still what your board supplier certifies. It is no longer sufficient evidence for furniture placed on the EU market, because REACH Annex XVII entry 77 sets a 0.062 mg/m³ limit on the finished article — half the E1 panel ceiling under the same chamber conditions.

What exactly has to be printed on a TSCA Title VI label?

Panels or bundles carry four fields: panel producer name, product lot number, compliance level (TSCA Title VI, or NAF/ULEF), and the EPA-recognised third-party certifier number. Finished goods or their boxes carry three: fabricator name, production date as month and year, and the compliance statement. The label can be a stamp, a tag or a sticker, and it has to be visible.

Does every piece in a flat-pack carton need its own label?

Under the federal rule the label goes on every finished good, or on every box or bundle containing finished goods. CARB's guidance is stricter in practice: when several separate finished goods share one package, each one needs its own label. Labelling both the item and the box costs nothing and settles the argument.

Where should the emission class actually appear so buyers stop asking?

On the image, next to the measured dimensions — because that is the asset that travels. A buyer forwards your product photo into a chat; they do not forward the certificate PDF. The practical fix is to lock the real, measured board thickness onto the product photo and put the compliance fields beside it: snap the dimension to the panel's actual edge, add the emission class and test method as labelled callouts, and export the same image at each marketplace's spec-image size. Software that measures the photo does this deterministically — the 18 mm on the drawing is the 18 mm that was measured, not a plausible-looking number an AI image generator produced. A worked example of that layout is here: annotated furniture spec image.

Sources & References

• US EPA — Formaldehyde Emission Standards for Composite Wood Products: scope, emission standards and compliance dates
• US EPA — Frequent Questions for Regulated Stakeholders: label contents for panels and finished goods, the 144-square-inch de minimis, three-year recordkeeping
• California Air Resources Board — Comparison of Key Requirements of CARB and U.S. EPA TSCA Title VI Regulations (updated January 2024)
• California Air Resources Board — Frequently Asked Questions: Labeling
• UL Solutions — New EU Restriction on Formaldehyde: Regulation (EU) 2023/1464, limits, dates and exemptions
• Bureau Veritas — REACH restriction on formaldehyde and formaldehyde releasers: Annex XVII entry 77 limits and test conditions
• SGS — Germany announces DIN EN 16516 as the reference method for formaldehyde emission, effective 1 January 2020
• GB/T 39600-2021 — Formaldehyde emission grading for wood-based panel and finishing products (standard catalogue entry)
• Zeleniuc & Coșereanu, RECENT vol. 17 (2016) — Standards and regulations concerning formaldehyde emission: EN 13986 class limits and JIS A 1460 desiccator correlations
• Japanese Standards Association — JIS A 1460:2015, determination of formaldehyde emission from building boards, desiccator method