Dresser Tip-Over Rules for Small Shops: ASTM F2057-23, the Three Tests, and Anchors

The rain comes back to Seattle in October, and for the next six months the shop smells like wet cedar no matter how tight the door seals. That's the season the kid's-room commissions arrive: a dresser for a nursery, a chest for a five-year-old who has outgrown closet bins. I've come to believe those are the highest-stakes pieces in the catalog, and not for the reason woodworkers usually worry about. Nobody is going to get hurt by a joint that opens up in year eight.

CPSC's 2023 Tip-Over Annual Report puts an estimated 6,400 emergency-department-treated furniture tip-over injuries to children on the board every year. Of those, 1,800, about 28 percent, involve a chest, bureau, or dresser. Across all tip-overs the agency counted an annual average of 17,800 ED-treated injuries for 2020 through 2022, and 217 reported fatalities from January 2013 through July 2023. Children under 18 accounted for 155 of those deaths, 71 percent of the total.

Since September 1, 2023, there has been a mandatory federal standard aimed squarely at that number, and it is written in units a shop can actually measure.

The Rule That Took Effect, and the One That Didn't

Two different CPSC rules share the citation 16 CFR part 1261, and mixing them up is the fastest way to get this wrong.

The one in force came out of STURDY, Pub. L. No. 117-328, Div. BB, tit. II, § 201, signed December 29, 2022. Per CPSC's own ballot vote sheet of March 22, 2023, STURDY told the Commission that if a qualifying voluntary standard was published within 60 days of enactment, by February 27, 2023, CPSC had to adopt it. ASTM published F2057-23 on February 6, inside the window. The Federal Register final rule of May 4, 2023 replaced part 1261 with that standard.

The other rule, published November 25, 2022, was CPSC's own test method, and it was stayed and never took effect. GAO confirms that in report B-334879. It matters because its numbers still circulate in compliance write-ups: per Intertek's bulletin on that version, a 57-pound loaded mass, 1.3 cubic feet of storage, a 1.5-degree average tilt, a hang tag carrying a stability rating, and a 105 percent anti-stockpiling cap. If a page you're reading mentions a hang tag, close it. The Federal Register rule says plainly that it does not require one.

Per Cornell's Legal Information Institute, 16 CFR § 1261.2 incorporates ASTM F2057-23, approved February 1, 2023, by reference, and points to free read-only access at astm.org/READINGLIBRARY. ASTM's store lists the standard at $104 for 22 pages, and the companion restraint spec F3096-23 at $64 for two, as of late August 2026. Read the free copy before you spend anything.

Three Numbers Decide Whether Your Piece Is Covered

CPSC's ballot sheet quotes ASTM F2057-23 § 1.1 directly. The standard covers free-standing clothing storage units (chests, chests of drawers, armoires, bureaus, door chests, dressers) that are 27 inches or greater in height, 30 pounds or greater in mass, and contain 3.2 cubic feet or greater of enclosed storage volume. The American Home Furnishings Alliance states it as an explicit AND. Miss any one of the three and the piece is out of scope.

The thresholds are not arbitrary. CPSC's ballot sheet records that the Commission found no known incidents causing death or serious injury for units weighing under 30 pounds empty, and that ASTM set 3.2 cubic feet from the lowest known volume of a fatal-incident unit 27 inches or taller.

Four scope details will bite a furniture maker before anything else does. The ASTM webinar deck hosted by AHFA in July 2023 lists "nightstand" as removed from the exemptions, so a tall, heavy nightstand is now judged on the numbers rather than its name. Intent governs generally: per that deck, what an item is named "may be a clue to intent, but it is not the only one," alongside how it's marketed and how consumers recognize it. Free-standing is broad, and the deck is explicit that an anti-tip device is not permanent attachment, so anchoring a piece to a wall does not take it out of scope. And levelers get adjusted to minimum before testing, to approximate the least a consumer will do.

Shelving, bookcases, office and dining furniture, jewelry armoires, underbed drawers, and built-ins are excluded under § 1.2.

Running the Three Tests on a Shop Floor

CPSC describes these as simulating a child weighing up to 60 pounds. Test 1 runs on a hard, level, flat surface.

Test 1, simulated clothing load (§ 9.2.1). Fill extendible elements, meaning drawers and pullout shelves, at a density of 8.5 pounds per cubic foot. Open everything. The unit must stay upright for 30 seconds. One wrinkle worth knowing: CPSC's ballot sheet phrases the loading conditionally, tied to whether 50 percent or more of the storage volume is extended, while AHFA and Intertek describe it as unconditional.

Test 2, simulated horizontal dynamic force (§ 9.2.2). With doors open and drawers extended, apply a 10-pound horizontal force over at least 5 seconds at the highest hand-hold, not exceeding 56 inches, then hold it at least 10 seconds. Eurofins gives the SI figure as 44.5 N at up to 1.42 m. The 10 pounds isn't a guess: CPSC cites a study putting the elbow pull strength of children 2 to 5 at 6.14 to 26.0 pounds.

Test 3, the carpet test (§ 9.2.3). Put a 0.43-inch test block under the rear legs, open all doors and extendible elements, and apply 60 pounds gradually to whichever element is most likely to cause tip-over, holding 30 seconds. Per CPSC, 60 pounds is roughly the 95th percentile weight of a 72-month-old on the 2000 CDC growth charts. F2057-23 raised that from 50 pounds and, more consequentially, applies it with all available drawers open rather than one.

Two apparatus notes from the ASTM deck are worth stealing. Exactly 60.0 pounds "is not possible," so a tolerance exists; and hardware counts as part of the test weight mass, which means your strap, hooks, and pins go on the scale.

If a piece fails, the deck lists what industry actually did: mass low and toward the back, shorter glide lengths, deeper cases, smaller drawer boxes, moving the fulcrum out as far as the design allows, lighter drawer fronts, and drawer interlocks. Most of those are free at the design stage and expensive after the case is glued up.

The Anchor Is the Part That Keeps Failing

F2057-23 requires an anti-tip device meeting ASTM F3096 to ship with every covered unit. Per Scott Buehrer of B. Walter & Co., presenting at the 2023 AHFA Summit, the F3096-23 test is simple: assemble the restraint, fix one end, attach a loading device to the other, ramp a 60-pound static load over 2 to 15 seconds, and hold it 30 more. Bureau Veritas confirms the timing and says the only substantive change from F3096-14 was raising that load from 50 to 60 pounds. ASTM dates the revision May 24, 2023; Bureau Veritas says June. Call it mid-2023.

Here is the limitation, in ASTM's own words on its store page: F3096 "assesses the strength of the tipover restraint only, and does not address the in situ performance." It tests the strap, not the strap plus your customer's wall.

Consumer Reports measured that gap. CR pulled 14 anchor kits to failure on an Instron, with test leader José Amézquita building roughly 90 wall segments. The Simple Mount kit averaged 347 pounds on wood studs. Ikea's Malm kit held 329 pounds on wood studs and failed at 58 on drywall. A Bassett Modern Rivoli kit popped at 23 pounds with drywall anchors. CR's headline: only 3 of 9 kits claiming drywall compatibility withstood 60 pounds.

The recalls follow the same fault line. Per Yahoo News, Cranach Hardware recalled 55,170 plastic tip restraint kits on October 24, 2025 after CPSC testing found they failed F3096-23, with 115 reported incidents. CR records a January 2024 action in which CPSC and 33 manufacturers recalled millions of New Age Industries zip-tie anchors that became brittle over time. AHFA reports F3096 is now under review, with CPSC failure data from 2013 through 2023 showing 16 anti-tip device failures, about half of them connection failures rather than device failures.

Watch what a package claims. Picture Hang Solutions lists a steel-cable Furniture Anti-Tip Kit, model S-HWR-ANTITIP, at $13.95 or $59.95 for ten as of August 25, 2026, in aircraft-grade galvanized steel, rated to "up to 400 lbs." That listing makes no ASTM F3096 claim. A load number is not a standards conformance statement, and for a covered unit you need the second thing.

Sale, Gift, and the Line the Statute Draws

There is no small-shop, custom-build, or unit-count exemption anywhere in 16 CFR part 1261 or in F2057-23's scope clause. So the question falls back on definitions. Per Cornell LII, 15 U.S.C. § 2052 defines a manufacturer as "any person who manufactures or imports a consumer product," and "manufactured" as "to manufacture, produce, or assemble." No volume floor appears in that text. A consumer product is one "produced or distributed (i) for sale to a consumer" for household use, "or (ii) for the personal use, consumption or enjoyment of a consumer." Craftybase, writing for handmade sellers, puts the trade reading bluntly: the moment you make a product for sale, you're a manufacturer in the eyes of the law. That's an interpretation, not an agency ruling, but it tracks the statute.

Read that against the piece on your bench. A dresser you build for your own bedroom was not produced or distributed for sale to a consumer. A commissioned dresser, or one listed on Etsy, was. That's the line the text draws, and I'd plan around it rather than hope for a carve-out.

If you sell one, the obligations are workable. Meet F2057-23. Ship an F3096 restraint. Apply the permanent warning label: per the ASTM deck, placement is 0.5 inch or less from the top of the drawer side and 3 inches or less from the back of the drawer front, or on a door or adjacent cabinet side no more than 1 inch from the closed position, with four placement options under § 10.1.1.1. Issue a General Certificate of Conformity under CPSA section 14(a)(1). AHFA is clear that third-party testing is not mandatory unless the product is marketed for use by children, and Compliance Gate notes a GCC can rest on first-party testing at any qualified facility. Your own shop test can support the certificate.

Market it for children, though, and everything changes: certification then runs under 14(a)(2) and (a)(3) with a CPSC-accepted lab meeting 16 CFR part 1112. Small batch registration does not rescue you there in the way people assume. It's relief from third-party testing for children's products, and it does nothing about meeting the standard, certifying, labeling, or shipping a restraint.

What I'd Do Before the Truck Leaves

Measure the three numbers first, at the design stage, on paper. If a piece lands within a few inches or a few pounds of the thresholds, decide deliberately which side of the line you want it on, and let the case dimensions follow that decision instead of habit.

Then build the test rig once. A flat slab, two 0.43-inch blocks, a 60-pound weight with symmetric handles, a 10-pound pull with a scale in line, and a stopwatch. That's an afternoon and maybe fifty dollars of steel, and it turns a legal question into a measurement you can repeat on every piece that leaves the shop.

And send the customer a restraint that would survive the wall it's going into. The standard tests your strap. It does not test their drywall, and Consumer Reports' drywall numbers say that's where these actually fail.