Buying rough lumber: what you actually pay for

A hardwood yard's price tag and a home center's price tag are answering two different questions, and buying rough stock as though it works like dimensional lumber is how a board-foot total that looked right on paper turns into a shortfall at the bench. Rough lumber is priced, tallied, and graded by a different logic than a surfaced 2x4 entirely — here's what the number on the tag is actually counting, and what's left of it once the board is usable.

The quarter system, not the nominal-name system

Hardwood is sold by the quarter: 4/4 (four-quarters, a rough 1" board), 5/4, 6/4, and 8/4 (a rough 2" board) are the common sizes, and the number is counted in quarter-inches off the saw, set before a board has been anywhere near a jointer or a planer. This is a genuinely different convention from a softwood "2x4," which names a size the board was cut toward and then surfaced past — see why a 2x4 isn't 2x4 for that side of the comparison. A rough 4/4 board really does start life close to a true 1", and a rough 8/4 board close to a true 2"; the gap between the name and what ends up in your hand shows up later, at the planer, not at the saw that cut it to rough size.

What the tally actually counts

A yard tallies board feet off the rough, as-sawn dimensions — thickness in quarters, width measured and typically rounded to the nearest inch per board, length as cut — and prices from that rough tally, before the board has been surfaced, jointed, or squared at all. Take a rough 8/4 black walnut board, yard-measured at 9" wide (rounded from its actual, slightly irregular edge) and 96" long: 2 x 9 x 96 ÷ 144 = 12 board feet, and at black walnut's typical $14 per board foot, that tallies to $168 before a single pass through a planer. Nothing about that number is wrong or inflated — it's an honest, standard tally of the rough material you're buying — but it is not the board-foot count of what you'll actually be building with once the board is dressed to usable stock.

What's left after milling

Surface that same 8/4 board on both faces (S2S, the common state hardwood ships in — edges typically left rough for the buyer to joint) and a rough 2" thickness commonly finishes around 1-13/16" (1.8125"), a loss of about 3/16" split across both faces. Joint one edge straight and rip the other to clean up wane or a crooked edge, and a realistic finished width lands around 8.5" rather than the tallied 9". Square both ends to remove checking and drying stress, and a realistic finished length is closer to 94" than a full 96". Run the finished numbers: 1.8125 x 8.5 x 94 ÷ 144 = 10.0569 board feet — a yield of 83.81% against the 12 board feet you paid for, and an effective cost of $16.71 per finished board foot, not the $14 on the tag, once the $168 purchase is measured against what actually became usable material.

The thickness loss itself is worth a closer look, because it doesn't scale the way it's easy to assume. A planer removes roughly the same absolute amount of material regardless of how thick the board started — call it 3/16" for a typical two-sided surfacing pass — which means the percentage lost shrinks as the nominal thickness grows: a 4/4 board loses 0.1875 ÷ 1 = 18.75% of its thickness, while an 8/4 board losing the same 3/16" loses only 0.1875 ÷ 2 = 9.375% — half the percentage, for double the starting material. This is the practical case for buying thicker rough stock when a project's finished thickness allows it: the planer takes a similar bite in inches either way, so a thicker board gives up a smaller share of what you paid for.

Grading: what the premium above "rough lumber" is actually buying

Hardwood grades — FAS, Select, and the common grades below it — describe how much clear, usable material a board is expected to yield from its worse face, not its overall quality as wood. A board graded FAS is expected to produce a high proportion of long, wide, defect-free cuttings from both faces; the common grades below it are expected to yield less clear material per board, priced accordingly, and are frequently the more sensible buy for parts that will be cut small anyway — a drawer side or a rail doesn't need the same long clear run a table leg or a door stile does. Hardwood lumber grading covers the grade definitions themselves; the buying decision that follows from them is straightforward once the definitions are clear: buy the grade the largest single part in the project actually needs, and let smaller parts come from a lower, cheaper grade rather than paying the top grade's premium across an entire order.

The one number on this site that isn't a citation

Every price used above, and every typicalCostPerBf figure behind a species page on this site, is a 2026 US retail estimate for rough 4/4 stock in hobbyist quantities — not a sourced, cited figure the way the shrinkage and density numbers on the same pages are. There is no single citable source for lumber pricing the way there is for USDA shrinkage data, because price genuinely varies by region, supplier, quantity, and month in a way density does not. Treat the dollar figures on this site, this post included, as a reasonable planning estimate for the shape of a cost comparison — walnut costs meaningfully more than poplar, 8/4 wastes less percentage-wise than 4/4 — and get an actual quote from wherever you're buying before finalizing a project budget. Presenting an unsourced estimate as though it were a citation would be a worse failure than just saying plainly that it isn't one.

Buying for the finished number, not the tag number

The practical habit that follows from all of the above: figure the board feet the project actually needs finished, then buy against a realistic yield — 80–85% is a reasonable planning range for surfaced, jointed, squared hardwood parts from typical FAS or Select stock, tighter if the project needs unusually long or wide clear sections. Buying exactly the tallied board-foot total a cut list calls for, with no allowance for what the planer and the jointer are about to remove, is how a project runs short a board late in a build rather than during the buying trip, when replacing one board is a quick fix instead of a special order. The board foot calculator handles the rough-tally side of this directly — run the finished dimensions once you know them, pad the total by your expected yield, and price the padded total against the real rough tally before you commit to a board.