Skip to main content

How Many Deck Boards Do I Need? Rows, Gaps and the Fencepost

A six-inch board is three different widths, and the gap between boards is part of what they cover.

By Mohamed Zakrya

Updated · 9 min read

Share
Counting material, certifying nothing How many deck boards do I need A “6 inch” board is three numbers, and the one that counts is the one with the gap in it. 1 · COVERAGE called 6.00 in measures 5.50 in covers 5.75 in Board plus one gap. Only this one divides. 2 · ROWS ceil((span + gap) ÷ (board + gap)) N rows carry N boards but N − 1 gaps, so the gap is added back first. 3 · BOARDS PER ROW ceil(run ÷ stock) Counted per row, never pooled: a butt joint has to land on a joist, so offcuts do not carry. 4 · THE COUNT 16 × 12 ft deck 26 rows 52 boards 13 joists 728 screws TWO FENCEPOSTS AND ONE SURPRISE The missing last gap On a 103.30 in span the formula gives 19 rows. Drop the added gap and you get 18, which reach 103.25 in. Short is short. The extra joist N spaces need N + 1 members, the same way a fence with ten panels needs eleven posts. Twelve spaces, thirteen joists. A wider gap, fewer boards The gap is covering, not waste. Over the same 12 ft span, 5/16 in gaps drop the count from 26 rows to 25. Set it for drainage anyway. This counts material. It certifies nothing. Joist and beam spans, header and post sizing, footing depth, ledger fasteners and connectors are absent here on purpose. They depend on species, grade, loading and local code, and a wrong span table is a collapse rather than a bad estimate.
A six-inch board is three different widths, and the gap between boards is part of what they cover.

A “6 inch” deck board has three widths. Its nominal width is 6.00 inches, its actual dressed width is commonly 5.50 inches, and its coverage width includes both the board and the planned gap. With a 1/4 inch gap, that board covers 5.75 inches across the deck.

Those numbers answer different questions. Nominal width identifies the product, actual width tells you what the board itself measures, and coverage determines the number of rows. Dividing by 6 inches orders too few boards. Dividing by 5.5 inches ignores the gaps and can order more rows than the deck needs.

The reliable sequence is to measure the deck, confirm the board’s actual width, add the intended gap, and round every partial row upward. The deck calculator follows that sequence while also counting stock boards, joists from your supplied spacing, and face-screw locations.

Measure the span and the run separately

A rectangular deck has two relevant directions. The span is the distance crossed by successive board rows. The run is the direction each individual board follows. If boards run along the 16 foot dimension of a 16 foot by 12 foot deck, the rows cross the 12 foot span.

These dimensions are not interchangeable. The span determines how many parallel rows fit across the deck. The run determines how many stock boards must be joined end to end within each row. Rotating the decking exchanges those jobs and can change both the stock-board count and the number of butt joints.

Measure the finished deck surface rather than relying on an early sketch. Record the span in inches for the row calculation, then record the run and available stock-board length in the same unit. Mixing feet and inches inside one division produces a result that can look plausible while being wrong.

The verified example uses a 16 foot run and a 12 foot span. Its boards run 16 feet, while its 26 rows accumulate across 12 feet. Keeping those directions labelled prevents the common mistake of using deck area alone, which cannot describe row layout or stock-length breaks.

Nominal width names the board but does not cover the deck

Nominal width is the trade name used to identify dimensional softwood. Actual width is the measurement after the lumber has been surfaced and dried to its standard dressed dimensions. That processing is why a board called 6 inches wide does not necessarily measure or cover 6 inches.

Nominal-versus-actual softwood dimensions are standardized through Voluntary Product Standard PS 20, the American Softwood Lumber Standard, published by the US Department of Commerce. The label remains useful for identifying material, but your row calculation needs the actual measurement of the decking you will install.

A 6 inch board has three widths One board, three widths Drawn at 56 px to the inch. Only the bottom number divides into the span. Nominal what it is called 6.00 in a name, not a measurement Actual what it measures 5.50 in dressed, under PS 20 Covers board plus its gap 5.50 + 0.25 gap = 5.75 in Divide the span by 5.75, not by 6 and not by 5.5. Six under-orders because the board is narrower than its name. Five and a half over-orders because the gap covers ground too.
A nominal 6 inch board can be 5.5 inches wide and cover 5.75 inches when paired with a 1/4 inch gap.

Coverage introduces the third number. A 5.5 inch board with a 1/4 inch gap forms a 5.75 inch repeating module. The board provides the solid portion, while the gap provides planned space before the next board. Together, they control the distance added by another row.

Confirm actual width from the material rather than converting every product from its name. Hardwood decking may use different dimensions and is often sold by board volume, so the board foot calculator can address that separate quantity. Row coverage still depends on measured face width and the intended gap.

Add the gap to find coverage

Module width is the actual board width plus one gap. For the example board, the calculation is direct.

module width = 5.50 in + 0.25 in
module width = 5.75 in

The module belongs in the row-count divisor because each additional row normally adds one board and one intervening gap. Using nominal width pretends the board is wider than it is. Using actual width alone pretends the gaps occupy no deck surface. Coverage represents the installed repeating pattern.

A wider gap therefore lowers the possible row count, which can feel backward at first. Every row contributes slightly more coverage when its board is paired with a wider gap. Because the final count is rounded upward, small gap changes may leave the result unchanged until the calculation crosses an integer boundary.

Across the same 12 foot span with 5.5 inch boards, the verified results are:

GapRowsBoards with two per row
1/8 in2652
1/4 in2652
5/16 in2550
3/8 in2550
A wider gap means fewer boards A wider gap buys you fewer boards A 12 ft span crossed by 5.5 in boards. The gap is part of the covering, so widening it lowers the count. GAP MODULE ROWS BOARDS 1/8 in 5.6250 in 26 52 1/4 in 5.7500 in 26 52 the page default a row drops out here 5/16 in 5.8125 in 25 50 3/8 in 5.8750 in 25 50 rows = ceil((144 + gap) ÷ (5.5 + gap)) The gap is not waste. It is covering. Set it for drainage, debris and how the boards will move, then read the count. Do not set it to save boards.
Increasing the gap increases each row’s coverage, so the rounded row count can fall at a threshold.

This comparison explains the arithmetic, but it does not select an appropriate gap for your material or conditions. Enter the gap established for your project, then let that value affect coverage. The calculator’s job is to count rows from a supplied layout decision, not make that decision for you.

For the 1/4 inch case, the engine divides the adjusted 12 foot span by the 5.75 inch module. The extra gap in the numerator is a fencepost correction, not an added gap beyond the deck.

rows = ceil((12 أ— 12 + 0.25) أ· 5.75)
rows = ceil(25.087)
rows = 26

The first fencepost correction protects the far edge

It is tempting to divide the span directly by module width. That shortcut treats every row as though a gap follows it, including the last row. A real layout with N boards has only N − 1 internal gaps because nothing follows the final board within the measured span.

The physical reach of N rows is therefore the combined width of N boards plus the combined width of N − 1 gaps. You need that reach to equal or exceed the measured span.

reach = N أ— board width + (N − 1) أ— gap

Solving that relationship for N moves one gap to the span side. The resulting expression is the exact row formula used by the engine.

N أ— board width + (N − 1) أ— gap ≥ span
N أ— (board width + gap) ≥ span + gap
N ≥ (span + gap) أ· (board width + gap)

That numerator correction matters only within a narrow interval near some rounding boundaries, which makes the error hard to notice during casual checking. Many dimensions produce the same rounded result with or without it. When a span lands inside that interval, however, the shortcut undercounts by one complete row.

A 103.30 inch span with 5.5 inch boards and 1/4 inch gaps demonstrates the difference. The corrected calculation produces slightly more than 18 modules, so rounding upward requires 19 rows.

rows = ceil((103.30 + 0.25) أ· 5.75)
rows = ceil(18.0087)
rows = 19

Dropping the added gap produces slightly less than 18 and incorrectly returns 18 rows. Checking the physical layout exposes the failure: 18 boards provide 99.00 inches, while 17 gaps provide 4.25 inches. Together they reach only 103.25 inches.

incorrect rows = ceil(103.30 أ· 5.75)
incorrect rows = 18

reach = 18 أ— 5.50 + 17 أ— 0.25
reach = 103.25 in
shortfall = 0.05 in
Four boards, three gaps Four boards, but only three gaps The last board has nothing after it, and that missing gap is the whole correction. no gap here so the formula adds one back before dividing 1 2 3 4 4 boards + 3 gaps N × board + (N − 1) × gap ≥ span N = ceil((span + gap) ÷ (board + gap)) Where dropping it changes the answer: a 103.30 in span with the gap added: ceil(103.55 ÷ 5.75) = 19 rows without it: ceil(103.30 ÷ 5.75) = 18 rows 18 rows reach 103.25 in. The deck stops 0.05 in short of the rim, and one row cannot be bought later to fix it.
Eighteen boards create only seventeen internal gaps, leaving the 103.30 inch span short by 0.05 inch.

A small shortfall remains a shortfall. The formula must test whether the installed boards and their actual internal gaps reach the rim, not whether an approximate module division nearly reaches it. Adding one gap before division accounts for the absent gap after the last board.

Count stock boards within each row

Rows describe coverage across the deck, but they do not yet tell you how many stock boards form each row. Divide the board run by the available stock-board length and round upward. A partial final length still requires another stock board for that row.

boards per row = ceil(deck length أ· stock board length)

For the verified example, each row runs 16 feet and the stock boards are 12 feet long. One board cannot complete that run, so every row requires two stock boards.

boards per row = ceil(16 أ· 12)
boards per row = ceil(1.333)
boards per row = 2

boards = 26 أ— 2
boards = 52

The engine does not carry a theoretical offcut from one row into the next. Each row’s final piece is bought whole because its butt joint must land on a joist, and the required cut position depends on that supported joint. Treating all row lengths as one continuous total can invent reuse that the framing layout cannot accept.

A detailed cut plan may reveal usable pieces when lengths, joint locations, and appearance constraints align. This row-based count does not decide that layout. It preserves the support condition by counting each row independently, giving the verified total of 52 boards before any separate project allowance.

Joists use a second fencepost correction

The joist count begins with deck length divided by the joist spacing you already have for the project. That division counts spaces, not framing members. Just as a fence with 10 panels needs 11 posts, N joist spaces require N + 1 joists.

joists = ceil((deck length in inches) أ· joist spacing) + 1

Using the example’s 16 inch on-centre input across a 16 foot length gives 12 spaces and 13 joists. The 16 inch value is an input to this verified example, not a spacing recommendation.

joists = ceil((16 أ— 12) أ· 16) + 1
joists = ceil(12) + 1
joists = 13

Two screws per board per joist is common practice for face-screwed decking, not a code requirement. On that counting assumption, each ordinary row-and-joist intersection contributes two screws. Every butt joint also places two board ends on one joist, so that location needs a second pair for the row.

With two boards per row, each row has one butt joint. The engine represents those extra intersections by adding boards per row minus one to the joist count. For 26 rows, 13 joists, and two boards per row, the verified total is 728 screws.

screws = 2 أ— rows أ— (joists + boards per row − 1)
screws = 2 أ— 26 أ— (13 + 2 − 1)
screws = 728

What this arithmetic does not certify

This guide counts decking rows, stock boards, joists from a spacing you supply, and screws from a stated face-screwing assumption. It does not design a deck, approve its structure, verify an existing frame, or certify that any input is suitable for construction.

It intentionally provides no joist spans, beam sizing, header sizing, post sizing, footing depth, ledger fastener schedule, or connector and hanger selection. Those decisions depend on species, grade, loading, construction details, site conditions, and local code. A wrong span table is a collapse rather than a bad estimate.

Take structural questions to your building department and lumber supplier. Give them the project conditions and follow the approved documents for your location. Their direction supplies the framing and connection requirements; the calculator can then count joists only after you enter the established spacing.

The same boundary applies to hardware and attachment details. The screw result is a quantity based on common face-screwing practice, not a hardware schedule or code determination. This guide counts material and certifies nothing, so do not use its arithmetic as structural approval.

Check the inputs before ordering

Confirm the finished span and run, the direction the decking will follow, the board’s measured face width, the intended gap, and the stock-board length. Keep each calculation in one unit, and make sure the joist spacing comes from the project information you were given.

Review the layout wherever rows meet borders, stairs, openings, posts, or changes in direction. Those features can divide the surface into separate counting zones. Calculate each zone from its own span and run instead of forcing an irregular deck into one large rectangle.

Finally, check the rounded results against the physical pattern: 26 rows across the example span, two stock boards in every row, 52 boards altogether, 13 joists from the supplied spacing, and 728 screws under the stated assumption. That check catches swapped dimensions and missing fencepost corrections before they become material shortages.

Questions people ask

Why does a board labelled 6 inches cover only 5.75 inches?

The label gives its 6 inch nominal width, not its installed coverage. The dressed board in this example measures 5.5 inches across its face, and the planned 1/4 inch gap creates a 5.75 inch repeating module. Coverage includes one board plus one gap, while the fencepost formula separately accounts for the missing gap after the final board.

Can I use the nominal width when the actual boards have not arrived?

You should use a confirmed actual face width rather than assume the nominal label equals the measurement. Product information may provide that dimension before delivery, and you can verify it against the material later. If the actual width changes, the module width and rounded row count can change too, especially when the result lies near an integer boundary.

Will increasing the gap always reduce the number of rows?

No, a wider gap lowers the row count only when the calculation crosses a rounding threshold. In the 12 foot example, both 1/8 inch and 1/4 inch gaps produce 26 rows, while 5/16 inch and 3/8 inch gaps produce 25. This arithmetic shows the count effect but cannot decide which gap is appropriate for your project.

Can I reuse the cutoff from one row in the next row?

You can reuse it only if a real cut plan places every resulting butt joint on a joist and satisfies the project’s other layout requirements. The calculator does not assume that compatibility. It counts each row’s final stock piece independently because pooling all offcuts can claim reuse at positions where the framing offers no supported joint.

How should I count boards around a picture-frame border?

Count the border and the field as separate layout zones. Their boards run in different directions, so each zone has its own span, run, actual width, gap, and stock-length divisions. This arithmetic cannot decide the final border geometry or joint arrangement. Establish that layout first, then calculate each rectangular portion and combine the resulting material counts.

What should I measure when replacing boards on an existing deck?

Measure the existing board’s actual face width, the gaps, each replacement length, and the locations where board ends are supported. Do not rely only on the nominal name stamped on a receipt or remembered from installation. Replacement work also requires checking the condition and suitability of the existing structure, which a board-count calculation cannot determine or certify.

Why does a butt joint increase the face-screw count?

A butt joint brings two board ends onto one supporting joist instead of one continuous board crossing that location. Under the calculator’s assumption, each end receives its own pair, adding one extra pair for every joint in each row. Two screws per board per joist is common practice for face-screwed decking, not a code requirement.

Does the row calculation include angled cuts and irregular edges?

No, the core row calculation models rectangular coverage using a constant span, board width, and gap. Angled edges, curves, openings, and direction changes can create unequal row lengths and offcuts that the rectangle cannot predict. Divide the drawing into measurable zones where practical, then review the resulting cuts against the actual layout before relying on the combined count.