Methodology: how these weights are computed
Every weight on this site is one function, applied consistently: convert a species' basic specific gravity to a density at a given moisture content, then multiply by the board's actual (not nominal) volume. This page writes out that function, checks it against the handbook's own worked example, and states plainly where it does and doesn't hold.
The model
Basic specific gravity (G_b) is ovendry mass over green
volume — the USDA Wood Handbook's "Green" row for each species. Below the
fiber saturation point (30% moisture content, the point past which wood's
volume stops changing with moisture), the fraction of a species' total
shrinkage that has already happened at moisture content M is:
a = (30 − M) / 30
and a = 0 at or above 30% MC. Specific gravity referenced to
that moisture content, Wood Handbook Eq. 4-13:
G_M = G_b / (1 − 0.265 · a · G_b)
Density, Eq. 4-14:
ρ = 62.43 · G_M · (1 + M / 100) lb/ft³
A board's weight is then ρ × actual volume × quantity, where
actual volume uses the dressed dimensions from PS 20-25, not the nominal
label on the tag. 62.43, not the more familiar 62.4: the handbook states that
constant explicitly, and the published density tables we check this model
against were computed with it — rounding to 62.4 introduces a small
systematic low bias against every number on this site.
Worked example
White ash, G_b = 0.55, at 12% MC. First a = (30 − 12) /
30 = 0.6. Then G_12 = 0.55 / (1 − 0.265 × 0.6 × 0.55) =
0.6027. Then ρ = 62.43 × 0.6027 × 1.12 = 42.1 lb/ft³ —
against the handbook's own published 42.2 lb/ft³ for white ash at 12% MC. That
0.1 lb/ft³ gap is representative of the whole model's error, below.
How accurate is it
Checked across the 46 curated species — computing 12%-MC specific gravity
from G_b through Eq. 4-13 and comparing to the handbook's own
published 12% row — the deviations are: mean absolute deviation 1.86%, median
1.59%, max 8.88% (American basswood), and 89% of species within 4%.
Some of that scatter belongs to the source, not the model: the handbook's Green and 12% rows are independent measurements on different physical samples, and they are not always perfectly mutually consistent to begin with. Real board-to-board variation within a species runs ±10-15%, which dwarfs all of this — the number on this page tells you where the middle of the pile is, not what any one board weighs.
Where the model stops
Eq. 4-12 estimates a species' total volumetric shrinkage (green to
ovendry) as a linear function of basic specific gravity. That linear estimate
breaks down above G_b ≈ 0.8708: past that point the predicted
volume gain from wetting outpaces the mass of water added, and computed
density would start falling as a board gets wetter, which is not physical.
Commercial softwoods and hardwoods carried on this site stay well under that
line — the densest is around 0.66 — but a handful of very dense tropical
species (lignum vitae, roughly 1.05) exceed it. Those species are excluded
from the site rather than published with a number the model cannot compute
honestly.
Moisture presets
8% is a typical indoor in-service equilibrium moisture content (Wood Handbook Table 4-2). 12% is the handbook's own reference condition — what its property tables are published at — and this site's default. 15% is typical air-dried stock sitting at a yard. 19% is the PS 20-25 legal ceiling for a KD/S-DRY grade stamp at the time of surfacing; it is a maximum, not a typical value. Green has no single number: it reads each species' own heartwood and sapwood moisture content from Wood Handbook Table 4-1, tables on this site show the full heart-to-sap range, and the calculator's single-number answer uses the midpoint of that range as the honest point estimate.
Green-moisture proxies
Table 4-1 does not cover every species this site carries. Where the handbook's own row is missing, a congener's row stands in, disclosed here rather than silently substituted:
- Interior Douglas Fir uses the coast-type Douglas Fir row — the same species, a different growing region, and the only proxy here that isn't really a proxy.
- Eastern White Pine uses Western White Pine, its closest congener in the same soft-pine group.
- White Spruce and Red Spruce both use Black Spruce.
- Red Maple uses Silver Maple, a fellow soft maple.
- Shagbark Hickory has no Table 4-1 row of its own; we average the mean of Mockernut and Pignut hickory, its tight congeners.
- Quaking Aspen and Eastern Cottonwood use the handbook's own genus-level rows.
- Black Locust and Northern White Cedar have no defensible stand-in. Green weights are simply not shown for them — an absence, not an invented number.
Dressed sizes
Actual dressed dimensions come from NIST PS 20-25 Table 3, which gives separate figures for dry (surfaced at ≤19% MC) and green (surfaced at >19% MC) lumber — green stock is cut oversized to allow for shrinkage during seasoning. The popular shorthand "nominal widths over 6 in. lose 3/4 in. instead of 1/2 in." is wrong at 7 in.: a nominal 7 in. width loses only 1/2 in. (6-1/2 in. actual); the break to 3/4 in. does not happen until nominal 8 in. (7-1/4 in. actual). Separately, PS 20 §6.2.5 / Table 5 carve out different, larger green sizes specifically for Redwood, Western Red Cedar, and Northern White Cedar. That carve-out is not modeled here — this site's green figures for those three species use the general table, which runs marginally conservative (slightly light, not heavy).
Board feet vs. weight
Board feet are computed from NOMINAL dimensions, because that is how lumber is priced: a 2×4×8 is 5.33 board feet from 2 × 4 × 8 ÷ 12, not 4.00 from its actual 1.5 × 3.5 in. cross-section. Weight uses ACTUAL dressed dimensions, because that is what a scale reads. Roughly half the lumber calculators online conflate the two and get board feet wrong as a result.
See data sources and provenance for where every column of underlying data comes from, or a live example of this model in use: 2×4 Douglas fir weight and every species in a 2×4.