Every other plywood calculator divides by 32
It is worth being specific, because this is the whole reason this page exists. Work through the tools that rank for this search and they all do the same thing — total area, divide by 32, round up. One of them states the method outright: "A standard 4×8 sheet covers 32 square feet. To estimate how many plywood or MDF sheets you need, divide your total project area by 32."
That is arithmetically true and practically wrong, because plywood is not a liquid. A sheet has to break on a joist or a rafter, and the offcut left at the end of a row is rarely the size the next row needs. So the real count is a grid — rows across the width, sheets along the length — and it only equals the area figure when both dimensions divide evenly.
We measured how often that happens. Sweeping every room from 8 to 40 feet in half-foot steps, 4,225 combinations in total, the area method comes up short on 89% of them, by 20% on average. The worst case is a doubling: a room just over 8 ft square is 3 sheets by area and 6 by layout, because you need two rows and two columns whatever the arithmetic says. A test in this repo re-runs that whole sweep and fails the build if those numbers ever stop being true.
The calculator shows both figures and the gap between them, so you can see when your room happens to be one of the tidy ones — a 16 × 24 ft floor genuinely is exactly 12 sheets — and when it is not.
Why 10% overage does not fix it
The obvious objection is that a waste factor already covers this. It does not, and the reason is that the two numbers are covering different things. Overage covers a panel that arrives damaged, a cut spoiled around a stairwell, a sheet dropped off a scaffold. The layout shortfall is not waste at all — it is coverage you never had.
More practically, a percentage is the wrong shape for the problem. The shortfall averages 20% and routinely runs past 30%, so a 10% factor undershoots the majority of rooms it is meant to protect. And on a room that divides perfectly the shortfall is zero, where 10% is pure surplus. On this page the overage is applied on top of the layout count, which is where a percentage genuinely belongs.
Reading the two numbers on the stamp
APA RATED SHEATHING carries a Span Rating printed as two numbers with a slash — 24/16, 32/16, 40/20, 48/24. They are not a fraction and they are not a size. The left number is the maximum support spacing when the panel is used on a roof; the right is the maximum when it is used as subfloor. APA's own example: a 32/16 panel is good for roof decking to 32 inches on centre, or subflooring to 16.
The asymmetry is real and it is large — the same sheet spans twice as far overhead as it does underfoot. Floors take concentrated loads and are held to a tighter deflection limit, so a panel that feels solid on a roof deck can be bouncy as a floor.
Two things on that stamp catch people out. First, a right-hand number of zero— as in 24/0 — means the panel has no subfloor rating at all, at any spacing. It is not "spans a little"; it is not rated. 24/0 and 24/16 sit beside each other on the rack and only one of them belongs on your floor. Second, at 24 inch centres a subfloor needs a48/24 panel: a 40/20 does not reach, even though it looks like the next size down and its roof number is comfortably larger than 24.
APA RATED STURD-I-FLOOR is a different product with a single number — 16, 20, 24, 32 or 48 oc — because it is a combined subfloor and underlayment for a single-layer floor. Reading a single number as though it were the left half of a sheathing rating is an easy and expensive mistake.
Three supports, and why orientation is not free
Buried in APA's definition, and stated twice, is a condition most summaries drop: the Span Rating applies with the panel's long dimension or strength axis "across three or more supports" — that is, spanning at least two bays. A panel bridging only two supports is outside the basis of its rating.
This is why turning sheets to save a panel is a framing decision rather than a purely economic one. This calculator does try both orientations and reports the cheaper grid, because on a non-square room one way round is often a sheet or two better. But on a structural floor or roof deck, check that the orientation you choose still runs the strength axis across the supports — otherwise the saving is not real.
The related pieces of the same job live on their own pages: the stud calculator counts the framing this sheathing lands on, the deck joist span calculator handles how far those joists may run in the first place, and the drywall calculator does the same sheet-goods arithmetic for the inside face of the wall.