Which Wood for a Cutting Board? The FDA Food Code's Two Tests: Janka Hardness and End Grain

We've all been there. The wedding is three weeks out, there's a good stack of offcuts on the bench, and somebody on a forum has just told you, with total confidence, that the walnut you planned to use is unsafe, that oak harbors bacteria, and that anything but maple is asking for trouble. Forty comments later, nobody has cited anything and you still have to pick a wood before the kids get home.

What surprised me when I finally read the rulebook is that it's shorter than the forum thread. The FDA Food Code names one species, gives you two properties to judge everything else by, and stops. Both are things you can check at your own bench with a reference table and a hand loupe.

What the Food Code Actually Says (It's One Sentence)

The 2022 FDA Food Code, section 4-101.17, opens by banning wood: "wood and wood wicker may not be used as a FOOD-CONTACT SURFACE," with a few exceptions. The one that matters to us is paragraph (B): "Hard maple or an equivalently hard, close-grained wood may be used for: (1) Cutting boards; cutting blocks; bakers' tables; and UTENSILS such as rolling pins, doughnut dowels, salad bowls, and chopsticks."

That's it. The word "maple" appears exactly once in the whole Code, right there. "Hard" is never defined. "Close-grained" is never defined. There is no species list and no hardness number. The 2009 edition used identical wording. One sentence has outlasted a lot of arguments.

Before you get anxious: the Food Code is a model code, offered for states to adopt. It binds a restaurant in Minnesota because Minnesota Rules 4626.0490 carries the wording into state law; it does not bind your kitchen. Wood needs an exception because the general rule for food-contact materials, 4-101.11, requires them to be "Durable, CORROSION-RESISTANT, and nonabsorbent," and wood is absorbent. And the stated reason is not what the forums assume. FDA's Annex 3 explains 4-101.17 in three sentences, and the one that counts is: "Moist foods may cause the wood surface to deteriorate and the surface may become difficult to clean." It's a cleanability and durability rule. Bacteria hiding in pores are never mentioned.

Test One: Is It as Hard as Hard Maple?

"Equivalently hard" has an obvious yardstick even if the Code doesn't name it. The USDA Forest Products Laboratory's Wood Handbook defines hardness as the load needed to push an 11.28 mm (0.444 in.) steel ball into the wood to half its diameter, the Janka test, with side hardness measured perpendicular to the grain on the plank face. Sugar maple, the trade's hard maple, comes in at 1,450 lbf in the Handbook's Table 5-3b at 12% moisture.

The Code won't say how close to 1,450 is close enough, so any cutoff is ours, not FDA's. The Wood Database's cutting-board guide offers a practical band: below about 1,000 lbf a board "will start to show signs of wear faster than most would deem acceptable," and around 2,000 lbf boards "cause knife blades to dull too quickly." One caution: the Handbook's Table 5-6 puts the coefficient of variation for side hardness at 20 percent, so treat any Janka number as a neighborhood, not an address.

That band is hard on the two most popular boards after maple. The Handbook lists black walnut at 1,010 lbf and black cherry at 950, roughly 30 to 35 percent softer than the Code's named species and right at the Wood Database's wear floor. No source I've read calls them "equivalently hard." The industry uses them anyway; John Boos sells cherry and walnut boards alongside maple. They're lovely boards. They will just show knife tracks sooner, worth saying out loud when you hand one over as a gift.

Here's the failure mode that catches busy people: "soft maple" isn't a species. The Wood Database explains it's a trade grouping of red, silver, box elder, striped and bigleaf maple, running 700 to 950 lbf, and hard maple is "roughly twice as hard." At The Wood Yard in Concord, Georgia, 4/4 hard maple is $8.35 a board foot while soft maple is $8.25, so the price tag won't warn you. Two checks from The Wood Database: under a 10x loupe, hard maple's end-grain rays come in two distinct widths while soft maple's are uniformly thin, and half a teaspoon of ferrous sulfate in half a cup of water turns hard maple "a paler bluish green" and red maple "a deep, inky blue/black." Check before the glue-up, not after.

Test Two: Read the End Grain with a Loupe

"Close-grained" is harder to pin down because it isn't an anatomical term; the Wood Handbook's chapter on wood structure never uses it. What it does describe is how hardwoods arrange their vessels, the pores on end grain, which typically run 50 to 200 micrometers across. In diffuse-porous woods the pores "do not markedly differ in size" from earlywood to latewood; maple, birch and yellow-poplar are the Handbook's examples, and its illustration of a diffuse-porous wood is sugar maple itself. In ring-porous woods the earlywood vessels are big and shrink "often by an order of magnitude or more" in the latewood, leaving a visible ring of large holes every year; ash, hickory and red oak are the examples. Black walnut sits between as semi-ring-porous, though Woodcraft's WoodSense column calls it ring-porous; even the references don't fully agree.

The bench version is simple. Sand a scrap of end grain and look at it with the 8x to 10x hand lens Woodcraft recommends. A ring of large pores around each growth ring means ring-porous, which fails the plain reading of "close-grained." Small pores scattered evenly means it passes. The Wood Database describes hard maple's vessels as "small to medium" and red oak's earlywood vessels as "very large."

Red oak gives you the most satisfying test of all. Its big pores run open through the wood, with tyloses "absent or scarce," and The Wood Database's guide to telling the oaks apart says that if you can blow air through a piece of end grain, "it's probably red oak." A board you can blow bubbles through is not close-grained by any definition.

White oak is the argued exception. It's ring-porous too, but its pores are packed with tyloses, the plugs that let coopers make barrels from it. Whether plugged pores count as "close-grained" is a judgment the Code doesn't make, and no source I've found rules on it. My reading: white oak makes a fine, durable board, but if you want to claim the letter of the Code, maple, beech and birch are safer.

The Scorecard: Who Passes Both Tests

Put the two checks together and the forum argument collapses into a short list. Hardness is from the Wood Handbook; end grain from the Handbook and The Wood Database.

Species	lbf	End grain	Verdict
Hard maple	1,450	diffuse-porous	passes, by definition
American beech	1,300	diffuse-porous	passes both
Yellow birch	1,260	diffuse-porous	passes both
White oak	1,360	ring-porous, pores plugged	hard enough; grain arguable
Red oak	1,290	ring-porous, pores open	hard enough; fails grain
White ash	1,320	ring-porous	hard enough; fails grain
Black walnut	1,010	semi-ring-porous	grain borderline; too soft
Black cherry	950	semi-ring to diffuse	passes grain; too soft
Teak	1,000	ring or semi-ring	fails both

Beech and yellow birch are the quiet winners: both within about 15 percent of maple, both diffuse-porous, and yellow birch is rarely mentioned in cutting-board articles. Hickory goes the other way: shagbark is the Handbook's ring-porous example, and the hickories run 1,820 to 2,140 lbf, past the knife-dulling line. Teak adds insult to injury: The Wood Database reports up to 1.4% silica with "a pronounced blunting effect on cutting edges."

What the Germ Studies Actually Found

If you've been picturing salmonella setting up house in oak pores, the research should let you breathe. Dean Cliver's group at UC Davis ran their experiments "with more than 10 species of hardwoods" and, in his words, "found essentially no differences among the tested wood species." The 2016 review by Aviat and colleagues in Comprehensive Reviews in Food Science and Food Safety lists the species in the 1994 Ak, Cliver and Kaspar boards: ash, basswood, beech, birch, butternut, cherry, hard maple, oak and black walnut, a spread from 410 to 1,450 lbf that includes ring-porous ash and oak. The review concludes that wood's porous nature "is not responsible for the limited hygiene of the material... and that it may even be an advantage for its microbiological status."

So the hardness-and-grain rule is about a board staying smooth and cleanable through years of wet use, exactly as Annex 3 says. The pore-size argument is really a durability argument. For the home kitchen, the University of Maine Extension's advice is enough: half a teaspoon of unscented bleach per quart of water, left on the board for a minute.

From My Workshop to Yours: What I'd Glue Up

With three projects going and dinner to think about, here's the short version. If the board is a gift and you want to say honestly that it meets the one written standard, use hard maple, American beech or yellow birch, and check the maple with the loupe or ferrous sulfate before you cut. Maple isn't even the expensive choice: on The Wood Yard's August 2026 list, 4/4 hard maple is $8.35 a board foot against $10.00 for white oak and $16.50 for walnut, and Advantage Lumber ships 8/4 hard maple at $6.88. Cherry at $6.75 is the cheapest species that passes the grain test and makes a beautiful board for someone who's fine with it wearing sooner. Red oak at $6.10 is the cheapest "hard enough" wood and the one you can blow bubbles through. Skip it.

Maple, beech, the oaks, walnut and teak are all on The Wood Database's list of dust sensitizers; that's a respirator matter for the shop, covered in our toxic-woods piece, not a food matter. What you owe the couple is a board hard enough not to groove, tight enough to stay smooth, and labeled with what it's made of. Go sand a scrap of end grain tonight and take a look. It takes less time than reading the forum thread.