A 2x4 measures 1-1/2 by 3-1/2 inches, and once you have heard it, it stops being surprising. The arithmetic behind it is worth its own page all the same: a dressed size is not a scaled-down nominal size. It is the nominal size minus a fixed amount of material — and a fixed amount is a larger share of a small board than of a large one.
The allowance is a length, not a percentage
| Size | Nominal (in) | Dressed (in) | Off the thickness (in) | Off the width (in) | Section kept |
|---|---|---|---|---|---|
| 1x2 | 1 × 2 | 0.75 × 1.5 | 0.25 | 0.5 | 56.3% |
| 1x3 | 1 × 3 | 0.75 × 2.5 | 0.25 | 0.5 | 62.5% |
| 1x4 | 1 × 4 | 0.75 × 3.5 | 0.25 | 0.5 | 65.6% |
| 1x6 | 1 × 6 | 0.75 × 5.5 | 0.25 | 0.5 | 68.8% |
| 1x8 | 1 × 8 | 0.75 × 7.25 | 0.25 | 0.75 | 68.0% |
| 1x10 | 1 × 10 | 0.75 × 9.25 | 0.25 | 0.75 | 69.4% |
| 1x12 | 1 × 12 | 0.75 × 11.25 | 0.25 | 0.75 | 70.3% |
| 2x2 | 2 × 2 | 1.5 × 1.5 | 0.5 | 0.5 | 56.3% |
| 2x3 | 2 × 3 | 1.5 × 2.5 | 0.5 | 0.5 | 62.5% |
| 2x4 | 2 × 4 | 1.5 × 3.5 | 0.5 | 0.5 | 65.6% |
| 2x6 | 2 × 6 | 1.5 × 5.5 | 0.5 | 0.5 | 68.8% |
| 2x8 | 2 × 8 | 1.5 × 7.25 | 0.5 | 0.75 | 68.0% |
| 2x10 | 2 × 10 | 1.5 × 9.25 | 0.5 | 0.75 | 69.4% |
| 2x12 | 2 × 12 | 1.5 × 11.25 | 0.5 | 0.75 | 70.3% |
| 4x4 | 4 × 4 | 3.5 × 3.5 | 0.5 | 0.5 | 76.6% |
| 4x6 | 4 × 6 | 3.5 × 5.5 | 0.5 | 0.5 | 80.2% |
| 6x6 | 6 × 6 | 5.5 × 5.5 | 0.5 | 0.5 | 84.0% |
| 8x8 | 8 × 8 | 7.25 × 7.25 | 0.75 | 0.75 | 82.1% |
Nothing in the two "off" columns is proportional. The thickness loses 0.25 in (7), 0.5 in (10), 0.75 in (1), where each pair is the amount taken off and how many of the 18 sizes take it; the width loses 0.5 in (11), 0.75 in (7). What leaves the board is a fixed strip, and it does not grow with the board.
That is why the last column falls as the section falls. A 1x2 keeps 56.3% of its nominal section and a 6x6 keeps 84.0%; the gap between them is 27.8%. It is geometry rather than care: the strip runs round the perimeter, so the fraction it takes is set by the perimeter divided by the area, and that ratio is worst on the smallest sections.
The column does not even rise steadily. The 1x8 keeps 68.0% where the 1x6 below it keeps 68.8%. The 2x8 keeps 68.0% where the 2x6 below it keeps 68.8%. The 8x8 keeps 82.1% where the 6x6 below it keeps 84.0%. Wider is a bigger board and it keeps a smaller fraction of itself, because 0.75 in is where the width allowance steps up — and on the 8x8 the thickness steps up with it.
The 1 in and 2 in ladders print identical percentages on all 7 widths they share. That is not rounding: a 1x4 and a 2x4 both lose a quarter of their thickness and the same 0.5 in off their width, so every ratio in the two rows is the same number.
Stiffness is the cube of the depth
| Size | Section kept | Stiffness kept, on edge | Stiffness kept, flat |
|---|---|---|---|
| 1x2 | 56.3% | 31.6% | 31.6% |
| 1x3 | 62.5% | 43.4% | 35.2% |
| 1x4 | 65.6% | 50.2% | 36.9% |
| 1x6 | 68.8% | 57.8% | 38.7% |
| 1x8 | 68.0% | 55.8% | 38.2% |
| 1x10 | 69.4% | 59.4% | 39.0% |
| 1x12 | 70.3% | 61.8% | 39.6% |
| 2x2 | 56.3% | 31.6% | 31.6% |
| 2x3 | 62.5% | 43.4% | 35.2% |
| 2x4 | 65.6% | 50.2% | 36.9% |
| 2x6 | 68.8% | 57.8% | 38.7% |
| 2x8 | 68.0% | 55.8% | 38.2% |
| 2x10 | 69.4% | 59.4% | 39.0% |
| 2x12 | 70.3% | 61.8% | 39.6% |
| 4x4 | 76.6% | 58.6% | 58.6% |
| 4x6 | 80.2% | 67.4% | 61.4% |
| 6x6 | 84.0% | 70.6% | 70.6% |
| 8x8 | 82.1% | 67.5% | 67.5% |
What governs deflection is not the area but the second moment of area, and that goes as the breadth times the cube of the depth. The depth is also the dimension that just lost a quarter or a half, so the cube compounds the loss instead of averaging it. On 18 of the 18 sizes, both orientations keep less stiffness than section: the loss is not shared evenly between the two numbers, it concentrates in the one that gets cubed.
Which orientation suffers less is decided by the two ratios, not by the aspect ratio. On 13 of the 18 sizes a board keeps more on edge than flat, because its width lost a smaller fraction than its thickness. On the other 5 — 1x2, 2x2, 4x4, 6x6, 8x8 — the two columns are identical, and those are exactly the sections whose width and thickness lose the same fraction: the square sizes, plus the 1x2, where a quarter off one inch and half an inch off two both come to 25%.
Take the 2x4. On edge it keeps 50.2% of the stiffness its nominal numbers imply; laid flat it keeps 36.9%. The flip between those two positions is worse than the label prepares you for — the nominal numbers predict a factor of 4.0, and the dressed section gives 5.4, because the depth drops from 3.5 in to 1.5 in and the ratio is squared.
The range runs from the 1x2 at 31.6% to the 6x6 at 70.6% — and even the best of them is 29.4% short of the stiffness its nominal numbers imply.
What the board-foot figure counts
Board feet are computed on the nominal section, while the volume and the weight on the size pages are computed on the dressed one. A 2x4 × 8 ft is 5.33 board feet by the first measure and 3.50 by the second; a 1x2 × 8 ft is 1.33 and 0.75.
The ratio between the two figures is the section kept, on every size and at every length, because the length appears on both sides of the division and cancels. So the 56.3% to 84.0% range of the first table is also the distance between what a board-foot count says and what the board contains.
None of that is a markup. The board foot is the trade's unit and it is defined on the nominal section, so prices quoted per board foot already carry the difference. What the ratio tells you is how far the unit sits from the material in the stack — and that distance is widest on the smallest boards, the ones a shelf or a jig is actually made from.
Green stock is dressed larger
10 of the 18 sizes carry a second dressed size, for stock sold unseasoned that will shrink to the dry size as it dries. On 9 of them the green size is 0.0625 in larger in both directions. Fixed amount again, so again it is worth most where the section is smallest: 6.0% more section on a 2x4, 1.7% on an 8x8.
The 2x2 is the one exception. Green and dry dress to the same size there, so it is the only section where mixing the two does not leave a step — and the only one where a green-versus-dry substitution costs nothing in fit. Everywhere else the step is real: the 2x4 page works through what it does when green plates sit under dry studs.
Where this stops
- Prices, grades and availability. None of them are in these numbers. A dressed size exists in several grades, and which of them a yard stocks varies by region and season.
- Strength. This page is geometry. What a section is made of decides how much load it carries before it fails, and that is the wood stiffness guide, where sections are compared species by species.
- Deflection under a real load. The second moments above are per unit load. Span, support condition and elastic modulus are the other half, and the beam calculator does that arithmetic on the dressed section of a real board.
- Defects. A knot, a check or a slope of grain is not in a section calculation, and two boards from the same pile can differ by more than the gap between their nominal and dressed numbers.
- Shrinkage past the dressed size. Only the green dressed size is tabulated. How far a species moves as it dries is not in this database for sawn stock.
- Timber milled to other allowances. These are the North American dressed sizes. Timber milled to a different allowance has different actual sizes and none of these percentages transfer.
The lumber size index lists all 18 sizes and their length pages, and the Sitka spruce entry is the reference species behind the headline weight on each of them.