The joist count is bays plus one, and it happens three times
Divide the deck length by the on-centre spacing and you have the number of bays — the gaps between joists. You do not have the number of joists. Each bay is closed by a joist at both ends, and neighbouring bays share the joist between them, so one is left over at the end of the frame. Fifteen bays means sixteen joists.
This is the off-by-one that has its own name in software, borrowed from exactly this kind of carpentry, and on a deck it bites three separate times. The joists across the length are the obvious one. The beam lines across the width are the second: a 12 ft width at a 8 ft maximum joist span is two bays, which is three beam lines, not two. The posts along each beam are the third. Get all three wrong and a 20 × 12 deck comes up one joist, one beam and several posts short — a shortfall that looks like nothing on paper and stops the job on a Saturday morning.
The last bay is also almost never a full one. Spacing is a target, not a divisor: 240 in at 16 in leaves a final bay that is whatever remains, and that remainder is a real measurement you mark on the rim board. The tool reports it rather than rounding it out of sight, because a framer who does not know the last bay is short discovers it with the decking already ripped to width.
The gap between boards is a dimension, not a detail
Deck boards are not laid tight. There is a gap between them — typically 3 to 6 mm, or about 1/8 to 1/4 in — and it does two jobs: it drains water off the surface instead of holding it against the end grain, and it gives the board somewhere to go when it moves.
That gap is part of the width each board occupies. A 5.5 in board with a 3/16 in gap takes up 5.6875 in of deck, so a 12 ft width needs 26 rows, not the 27 you get from dividing by the bare board width. Ignoring the gap overestimates the board count on every deck you price, which is the quiet way a calculator sends you home with lumber you cannot return.
There is a second-order detail that most tools also miss: N boards have N−1 gaps, not N. The last board meets the edge and carries no gap after it. On a wide deck that costs you a fraction of a board; on a narrow landing it is the difference between finishing on a full board and finishing on a two-inch rip. The board count here is the smallest N where N × width + (N−1) × gap covers the deck.
On composite the gap is not advisory. Wood shrinks as it dries and gains clearance over time; composite does the opposite — it expands in heat, along its length as well as its width. A composite deck laid tight buckles in its first summer, and the manufacturer will point at the installation instructions rather than the warranty. That is why the composite preset here opens with a wider gap and with joists at 12 in on centre instead of 16 in.
Cut waste belongs on its own line
Commercial board lengths rarely land on the deck dimension. A 16 ft board on a 20 ft deck leaves a 4 ft butt joint that has to fall on a joist, and the offcut from the run is often too short to use anywhere. Add mitres, stair treads, picture-frame borders and the two or three boards that arrive cupped, and 10% is the normal allowance — closer to 15% on a deck with angles or a diagonal pattern.
The important part is not the number, it is where it appears. This tool applies waste as a percentage on top of the correct quantity and shows it as its own row, in the shopping list and again in the cost breakdown. Waste folded into a total is the reason two calculators disagree by 12% and nobody can say why. Shown separately, you can argue with it: set it to zero if you are buying exact lengths, push it to 15% if the deck has a diagonal deck pattern, and see immediately what that decision costs.
Labour is half the deck, and it is the half that varies
On a professionally built deck, labour is typically 40 to 60% of the total. That alone would justify splitting it out. What makes it essential is the variance: material prices are close to identical between two yards in the same town, and day rates are not. Two contractors can be 40% apart on the same drawing without either of them being unreasonable.
A single bottom-line number cannot be compared to a bid. A split can. Take the material subtotal here, compare it against the material line in the quote, and you learn something specific: if the bid's material is well under the quantities on this page, someone is planning to build with less wood — wider joist spacing, fewer posts, a thinner board, or a footing that is a paving slab rather than a poured pier. That is a conversation worth having before signing, not after the frame is up.
The tool offers labour two ways because contractors quote it two ways: as a rate per square foot of deck, or as a percentage of material. The rate-per-area mode adds railing labour per linear foot and stair labour per step separately, because those are the two parts of a deck where the hours are wildly out of proportion to the surface area.
A raised deck is a different deck
Cost per square foot is the number everyone searches for, and it is misleading the moment the deck leaves the ground. The surface above scales with area; the structure below does not.
| Height | Post spacing | What else changes |
|---|---|---|
| Under 30 in | ~8 ft | No guard required; labour is kneeling work |
| 30 in to 4 ft | ~8 ft | Guard becomes mandatory — a per-linear-foot cost the low deck never had |
| 4 ft to 8 ft | ~6 ft | More posts and footings; bracing; labour rises with staging |
| Over 8 ft | ~5 ft | Posts, footings and bracing again; two people to lift each piece |
Taller posts are more slender, and slender columns buckle before they crush, so the grid tightens as the deck rises. Every extra post is also an extra footing, dug and poured. The tool tightens the post grid automatically from the height you enter and multiplies the labour rate by a height factor, so the per-square-foot figure it reports moves the way real bids move rather than staying flat.
The 30 in threshold deserves particular attention because it is a cliff, not a slope. Below it, no guard. Above it, a guard around every open side — and railing is the most expensive linear foot on a deck, frequently more per foot than the decking is per square foot. Two decks with identical plans, one 28 in high and one 32 in high, are not within a few percent of each other.
Composite versus wood is a ten-year question
Pressure-treated pine is the cheapest way to get a deck. Cedar and redwood cost more and look better. Composite and PVC cost two to three times the material of pressure-treated per square foot, and want tighter joist spacing on top of that, which is why almost every side-by-side comparison ends there and declares wood the winner.
That comparison stops one line too early. Pressure-treated and cedar decks want cleaning and refinishing roughly every two years — five treatments over a decade, each of them product plus a weekend or a contractor's day across the whole deck area. Composite wants a wash. Enter your treatment cost in the tool and the ten-year total is calculated for both; the crossover typically lands somewhere between year six and year nine, earlier on a large deck than a small one because maintenance scales with area while the initial price gap is also per area but paid once.
None of which makes composite automatically right. It gets hot underfoot in direct sun, it cannot be sanded back when it scratches, and a hardwood deck genuinely looks better at year fifteen if it has been oiled. But those are aesthetic and comfort arguments, and they should be made on their own terms rather than smuggled in behind a price comparison that stopped at the checkout.
The two things the calculator cannot do for you
The ledger board. On an attached deck, the ledger is the board bolted to the house that carries half the deck's load. It is also, by a wide margin, the most common structural failure on a residential deck — not because the wood breaks, but because it was nailed instead of bolted, or fastened through a rim joist with no blocking, or installed without flashing so the wall behind it rotted quietly for eight years. This tool counts your material; it does not design that connection. Get it inspected.
Spans and sizes. The calculator counts how many joists you need at the spacing you enter. It does not tell you whether a 2×8 is deep enough for that span in that species and grade, or how big the footings need to be for your soil and frost depth. Those come from a span table or an engineer. Get the sizes first, then come back and count.
Both of these are also what the permit is for. In most jurisdictions an attached deck, or one above roughly 30 in, needs one — and the inspection exists precisely to catch the ledger and the footings. Check with your building department before ordering, because a rejected detail changes the framing plan and therefore every quantity on this page.
Privacy
Every figure stays in your browser. The arithmetic runs locally, nothing is uploaded, nothing is logged and there is no account. That matters more here than on most calculators, because what you type is your supplier pricing and, if you are the one bidding, your labour rate.
Frequently asked questions
How many joists does a 20 ft deck need at 16 in on centre?
Sixteen, not fifteen. 20 ft is 240 in, and 240 ÷ 16 is 15 — but 15 is the number of bays, the spaces between joists, not the number of joists. Every bay is closed by a joist at each end and neighbouring bays share the joist between them, so one is left over at the end of the frame: 15 bays, 16 joists. Dividing and stopping there removes exactly one joist, and it is always the one at the outer edge, which is where the decking has nothing under it and flexes when you walk on it.
Does the gap between deck boards change the number of boards?
Yes, and ignoring it makes you order too many. A 5.5 in board with a 3/16 in gap occupies 5.6875 in of deck width, so a 12 ft deck takes 26 rows rather than the 27 you get from 144 ÷ 5.5. The subtlety is that N boards have N−1 gaps rather than N — the last board meets the edge and carries no gap after it. On a wide deck the difference is a board or two; on a narrow one it is the difference between a clean edge and a ripped strip.
What percentage of deck cost is labour?
Usually 40 to 60% on a professionally built deck, and higher on a raised one. That is why this calculator shows material and labour as two separate subtotals with the labour share as a percentage. A single bottom-line figure cannot be compared against a contractor's bid; a split can. If a bid's material column comes in far under the quantities here, it is planning to build with less wood than the deck needs — wider joist spacing, fewer posts, or a thinner board.
Is composite decking cheaper than pressure-treated wood?
Not on build day, and usually yes by year ten. Composite runs roughly two to three times the material cost per square foot and wants joists at 12 in rather than 16 in on centre, which adds framing on top of the board price. Against that it needs no sanding, no stain and no oil. A pressure-treated or cedar deck wants refinishing about every two years, which over a decade is five treatments in product and labour across the whole deck area. Set the treatment cost in the tool and watch which column wins at year ten — the crossover is usually somewhere between years six and nine.
Why does a raised deck cost more per square foot than a ground-level one?
Because the structure underneath does not scale with the surface above. Taller posts are more slender and buckle sooner, so the post grid tightens — 8 ft spacing at ground level, closer to 6 ft above about 4 ft of height and 5 ft above 8 ft — which multiplies posts and footings. Labour per square foot rises too, because height means ladders, staging and two people to lift each piece. And above 30 in a guard becomes mandatory, adding a cost charged per linear foot that a ground-level deck does not carry at all.
Do I need a permit to build a deck?
In most US jurisdictions, yes, once the deck is attached to the house or is raised above a threshold height that is commonly 30 in. Freestanding low platforms under a certain area are often exempt. The permit is not paperwork for its own sake: the inspection catches the two failures that actually hurt people, which are the ledger connection to the house and the footing depth. Check with your building department before ordering material, because a rejected ledger detail can change the framing plan and therefore the quantities on this page.