'Food Safe' Cutting Board Finish: What FDA Rules Cover, and What They Don't

The board is done. Oak, glued up on a Tuesday between school pickup and dinner, sanded to 220, edges eased so it feels good in the hand. It's a wedding present, the wedding is in two weeks, and all that's left is the finish. Which is how you end up in the aisle reading cans that say food safe, salad bowl finish, FDA approved, none of which agree about what you should do next.

We've all been there, usually with three projects going and someone asking what's for dinner. So I did the slow version once and read the regulations finish makers point to. Some of the common advice holds up better than I expected. Some of it cites sections that don't say what people think they say.

The Phrase Isn't in the Rulebook

"Food safe" is not a category in the FDA regulations that finish literature cites. What exists instead is a framework for indirect food additives, substances that might migrate into food from something the food touches.

The general obligation is 21 CFR 174.5, which caps any such substance at the amount "not more than reasonably required to accomplish the intended physical or technical effect in the food-contact article." Read that last phrase again. Compliance is a property of the finished article in use, not of a can on a shelf. There's no test a can passes.

"FDA approved" on a finish can mean several different things. The ingredients might appear on a positive list. The supplier might hold a Food Contact Notification. A Threshold of Regulation exemption might apply. Or a law firm may have written an opinion that the substance isn't reasonably expected to become a component of food at all. The ACC notes that since 2000 the FCN has been the main route, and that FDA "generally will no longer issue or amend food additive regulations." Under 21 CFR 170.100(a) an FCN covers only the substance made by the manufacturer named in it, so clearance doesn't travel from one company's can to another's.

None of those routes is FDA testing a can of cutting board finish. That isn't a scandal. It just isn't what the label implies.

The Food Code Cares About Your Wood, Not Your Can

The one place where regulators do address cutting boards directly was the surprise of the whole exercise.

FDA Food Code 2022, section 4-101.17, bars wood as a food-contact surface and then carves out an exception: "hard maple or an equivalently hard, close-grained wood" may be used for cutting boards, cutting blocks, bakers' tables, rolling pins, salad bowls and chopsticks. That's the entire hook, species hardness and grain closure. The section says nothing whatsoever about applying a finish.

Section 4-501.12 adds the upkeep: cutting surfaces subject to scratching and scoring must be resurfaced when they can no longer be effectively cleaned and sanitized, or discarded if they can't be. The rule that actually governs your board is about what it's made of and whether you maintain it.

Film Finishes and the Barrier That Doesn't Last

The umbrella section people cite for coatings is 21 CFR 175.300. It covers a coating applied as a continuous film over a metal substrate, or one intended for repeated food-contact use applied to any suitable substrate as a continuous film "that serves as a functional barrier." Wood isn't mentioned anywhere in the section, but that second clause is the door a board finish would have to walk through, and it isn't obviously closed. Tung, linseed and soybean oil are on the permitted list, along with polyester and alkyd-type resins.

Two things undercut it. The first is polyurethane. The section that does name polyurethane resins, 21 CFR 177.1680, clears them as a food-contact surface for "bulk quantities of dry food," about as far from a wet, fatty, repeatedly cut board as a kitchen gets. And the ACC's list of CFR sections applicable to polyurethanes in food contact runs to nine entries and doesn't include 175.300 at all. I won't tell you polyurethane is banned on cutting boards, because nobody I could read says that. I'll say the clearance people assume exists is harder to find than the confidence around it suggests.

The second is the barrier premise, which fails on its own terms. A functional barrier has to stay continuous, and the Food Code's resurfacing rule assumes you'll eventually sand the surface off. Your knife starts that job the first afternoon.

General Finishes seems to know it. Their own page for Wood Bowl Finish makes no FDA or CFR claim at all, gives 30 days to cure, recommends the product for bowls, turnings and serving boards, then says plainly: "For countertops actively used for chopping and cutting, use Butcher Block Oil instead." The maker is more cautious than the folklore attributed to it.

Where the Rules Get Specific and Useful

Mineral oil is the happy exception, and 21 CFR 178.3620 sets three tiers. Paragraph (a), white mineral oil meeting the direct-additive spec at 21 CFR 172.878, may be used in nonfood articles and directly in or on food. Paragraph (b), technical white mineral oil, allows ultraviolet absorbance up to 4.0 at 280 to 289 millimicrons. Paragraph (c), refined mineral oil, tightens that to 0.7. Neither (b) nor (c) is cleared for direct food contact. Only (a) is.

Those absorbance limits are a proxy for what survives incomplete refining. You can't check absorbance in the shop, but you can check one word: USP. Howard's Cutting Board Oil page describes "100% Pure USP Food Grade Mineral Oil" and cites 21 CFR 172.878 and 178.3620(a) by section number, which is exactly the right pair. Oil sold for shop use with no grade stated isn't covered by paragraph (a), and there's no way to tell from the bottle.

Beeswax is cleaner still. Under 21 CFR 184.1973 it's affirmed GRAS for direct addition to food as a lubricant and surface-finishing agent. A board butter of USP mineral oil and beeswax is, uniquely among the things we put on boards, made entirely of substances with explicit direct-food clearances.

Notice the pattern: the products making the most modest chemical claim cite the most specific regulations. Tried & True says its blends are "completely 100% safe for skin and food contact," citing no CFR section; that rests on the company's own FAQ. Mahoney's Utility Finish, $9.99 for four ounces of heat-treated walnut oil, claims "completely food safe" and cites nothing either.

What "Safe Once Cured" Actually Covers

Bob Flexner's argument, in "The Folly of Food-Safe Finishes," is that the FDA list "includes every oil, resin, drier and additive commonly used in wood finishes" and "does not include lead or mercury," with a rule of thumb of 30 days to full cure.

On driers he's right, and you can check it in two browser tabs. Section 175.300(b)(3)(xxii) permits metallic driers built from aluminum, calcium, cerium, cobalt, iron, lithium, magnesium, manganese, zinc and zirconium, in salt forms including octoate (the 2-ethylhexoate) and naphthenate. Lead is not on that list. Then open the Minwax Tung Oil Finish Safety Data Sheet, which discloses cobalt 2-ethylhexanoate at 0.3% or less by weight. That's precisely the combination the regulation permits. The drier in a hardware-store oil finish is on FDA's own list for food-contact coatings.

Where it gets loose is "once cured." The 30-day figure appears in Flexner and on General Finishes' page, and I couldn't trace it to a test method or standard anywhere. Meanwhile the same Minwax sheet classifies the uncured product as Carcinogenicity Category 2 and Toxic to Reproduction Category 1B, with the statements "Suspected of causing cancer" and "May damage fertility or the unborn child." That's the wet product in your hand, and a strong argument for gloves and an open window. It isn't a verdict on the cured board. Nothing I could find measures how much of that cobalt stays available to migrate out of a cured film.

Flexner is more careful than the shorthand built on him. He concedes there "could also be problems with mineral oil and walnut oil that we just don't know of yet," and even with salad bowl finish, since none of it has been through FDA regulatory testing. That's not "film finishes are proven safe." It's "none of this is tested, and there's no evidence of harm." Repetition collapsed two different sentences into one.

From My Workshop to Yours

Here's where I land, and it's mine, not the government's. Anything that will meet a knife gets USP mineral oil, or mineral oil and beeswax. Flood it, let it soak, wipe the excess. Serving boards and bowls that never meet a blade are a different object, and a film finish there is a reasonable choice; cure it properly and stop worrying.

Then send the maintenance along with the gift, because that's the part the regulations care about. University of Maine Cooperative Extension's Bulletin #3108 recommends oiling with mineral oil or beeswax about once a month, and warns off olive and vegetable oil, which may become sticky and turn rancid over time. For sanitizing it gives half a tablespoon of unscented bleach per gallon of clean water, around 100 ppm available chlorine, one minute of contact, then wipe dry.

One caveat I won't paper over: the substrate research doesn't touch finishes at all. Dean Cliver's UC Davis lab found knife-scarred wood "acted almost the same as new wood" while scarred plastic was "impossible to clean and disinfect manually," and that same Maine bulletin still recommends plastic for raw animal products. I couldn't find a study testing whether oiling a board changes any of that.

So write the card. USP mineral oil, once a month. That instruction is worth more to the couple than any of the words printed on the can.