Deck Framing Spans 101: How the Load Gets to the Ground

The load path from decking to soil, what the IRC R507 span tables assume, how species and spacing change spans, and where prescriptive deck design ends.


Updated August 9, 2026

Every deck is a chain of handoffs

Before any span table makes sense, trace where the weight goes. A person standing on a deck loads a decking board; the board hands the load to the joists; the joists carry it to a beam (and, on an attached deck, to the ledger bolted to the house); the beam hands it down through posts; and each post presses its share into a concrete footing, which spreads it into the soil. Design a deck and you are really designing that chain, link by link — and the span tables in chapter R507 of the International Residential Code exist to size each link without an engineer.

The chain framing matters because each link only knows about the link above it: a joist cares only how much deck area it carries, a beam only how much deck leans on it, a footing only about the pounds arriving through its post and the soil underneath. Every table lookup below is one of these handoffs made explicit.

What the joist table actually assumes

The workhorse is 2021 IRC Table R507.6 — the same table published as Table 2 of the American Wood Council’s Prescriptive Residential Wood Deck Construction Guide (DCA 6). Its numbers encode a specific design condition: a 40 psf live load and 10 psf dead load, deflection limited to L/360 of the span, No. 2 grade lumber, and wet service conditions. Change an assumption and the table no longer applies — which is why decks under more than 40 psf of ground snow, or carrying concentrated loads such as hot tubs, need engineered design rather than a bigger row in the same table.

The span itself is measured between the faces of the supports — from the face of the ledger to the face of the beam — and it does not include any cantilever hanging past the beam. The cantilever gets its own rule: R507.6 caps the overhang at one-fourth of the joist span, so joists spanning 12 ft may run up to 3 ft past the beam — a quiet generosity that extends a deck past its beam line with no change to the framing.

Species and spacing move the numbers more than intuition says

Table R507.6 is organized by three species groups — southern pine; Douglas fir-larch, hem-fir, and spruce-pine-fir; and the redwood, western cedars, ponderosa pine, and red pine group — and the differences are not rounding error. A southern pine 2x8 at 16 inches on center spans 11 ft 10 in; the Douglas fir-larch group manages 11 ft 1 in; the redwood/cedar group less again. Match the group to the species stamped on the lumber, not to the treatment: pressure-treated stock in the Southeast is usually southern yellow pine; in the West, commonly Douglas fir or hem-fir.

Spacing is the other lever, and it is stronger than most first-time builders expect. That same southern pine 2x8 spans 13 ft 1 in at 12 inches on center but only 9 ft 8 in at 24 — tightening from 16 to 12 inches buys roughly 10 percent more span from the identical member. Spacing also interacts with the decking above: 24-inch spacing rules out most composite decking, and diagonal decking often forces 12-inch spacing regardless of what the joists could structurally span. Joist sizing is really a three-way trade among member size, spacing, and decking.

Beams: where the joists hand off

The beam table, 2021 IRC Table R507.5(1), answers the next question in the chain: how far can a built-up beam run between posts while carrying those joists? Its logic runs opposite to intuition: the longer the joists span, the shorter the beam may span, because each foot of beam now carries more deck. The table’s columns are joist spans at 6, 8, 10, 12, 14, 16, and 18 ft, and R507.5 says to round your actual joist span up to the next column: a deck with 11 ft joists is sized from the 12 ft column. Interpolation between columns is permitted by the table’s footnote; extrapolation past 18 ft never is.

As a feel for the magnitudes: a southern pine (2) 2x10 beam supporting 12 ft joists may span 7 ft 4 in between posts. Need more? Add a third ply or step up to 2x12s. Beams get a cantilever allowance of their own — footnote f permits the beam to overhang its end post by up to one-fourth of the adjacent beam span — mirroring the joist rule one level down the chain.

Posts and footings: pounds against soil

The last handoff is arithmetic rather than a table. Each post carries a tributary slice of the deck: half the joist span (the other half goes to the ledger) plus the full cantilever, multiplied by the post spacing, at the combined 50 psf design load. A deck with 12 ft joists, no cantilever, and posts 6 ft apart puts 6 × 6 × 50 = 1,800 lb on each post. The footing must spread that into the ground without exceeding what the soil can bear, and IRC Table R401.4.1 presumes 1500 psf for clay and silt mixes, 2000 psf for sand, and 3000 psf for gravel — with 1500 the required default when the soil is unclassified.

That 1,800 lb post on default 1500 psf soil needs 1.2 square feet of bearing — a 15-inch round footing, or 13 inches if a soils report justifies 2000 psf. Footings must also extend below local frost depth per R507.3.3 — and note that stronger soil shrinks every footing on the deck at once.

The ledger: half the deck hangs on it

On an attached deck, the ledger carries half of everything — and the code details this connection more closely than any joist or beam. The fastener schedule comes from IRC Table R507.9.1.3(1): half-inch lag screws or thru-bolts, spaced by joist span, staggered between a top and bottom row, each fastener 2 to 5 inches from the ledger ends, 2 inches down from the top edge, and 3/4 inch up from the bottom. The two fastener types are not interchangeable in quantity: at a 12 ft joist span, lag screws go at 15 inches on center where thru-bolts are allowed 29 — roughly twice as many lags for the same ledger.

Two more requirements matter more than the bolt count: R507.2.4 requires corrosion-resistant flashing over the ledger, because most ledger failures start as rot in the house band joist rather than bolt failure, and R507.9.2 requires a lateral load connection into the house — hold-downs rated at 1,500 lb at two locations, or 750 lb at four. The table also has a hard boundary: it covers a wood ledger fastened through up to 15/32-inch sheathing into a 2-inch nominal solid-sawn band joist, and attachment to or through brick veneer, stone, or hollow masonry is prohibited outright. A deck against a brick house must be freestanding or engineered.

Where the tables stop and an engineer starts

Prescriptive design is a fenced territory, and everything inside the fence has been solved for you. Outside it: joist spans past the beam table’s 18 ft columns, ground snow loads above 40 psf, concentrated loads like hot tubs, ledgers on masonry or engineered rim boards, and suspect fill — though merely unclassified soil is fine at the 1500 psf default. None of these is an "add 20% and hope" situation — the tables simply no longer describe your deck.

Inside the fence, the sequence is mechanical: size joists from R507.6, round the joist span up a column for the beam table, work the post loads down to footings against the soil, and fasten the ledger per R507.9 with flashing and a lateral connection. Our deck joist span, beam and footing, and ledger fastener calculators each run one link of that chain, tables included.

Frequently Asked Questions

Where exactly is a joist span measured from?

Between the faces of the supports: from the face of the ledger board to the face of the beam. The cantilever — the part of the joist hanging past the beam — is not included in the span and is limited separately to one-fourth of the joist span under R507.6. Your measured span must come in at or under the table value for the species group, size, and spacing you chose.

Why does the beam table ask for my joist span?

Because the joist span sets how much deck each foot of beam carries: longer joists mean a wider tributary strip of deck leaning on the beam, so the allowable beam span shrinks. Table R507.5(1) has columns at 6 through 18 ft of joist span, and your actual span rounds up to the next column — 11 ft joists are sized from the 12 ft column. Past 18 ft there is no column to round to; that deck needs an engineered beam.

Is a freestanding deck easier to get right than a ledger deck?

It trades problems. Freestanding decks skip the ledger entirely — no fastener schedule, no flashing detail, no lateral hold-downs, and it is the required route where a ledger is prohibited, such as attachment over brick veneer. The cost is a second beam line with its own posts and footings where the ledger would have been. On houses where the band joist is sound, accessible solid-sawn lumber, a properly fastened and flashed ledger is a well-mapped, code-tabulated connection.

When do I actually need an engineer for a deck?

Whenever your deck steps outside the tables’ stated assumptions: a hot tub or other concentrated load, ground snow above 40 psf, joist spans beyond the beam table’s 18 ft columns, a ledger that would land on brick veneer or hollow masonry, or an engineered rim board whose fastening the prescriptive table does not cover. The tables are not padded enough to absorb these by guesswork — they simply stop describing your structure.

Try the Calculators

Sources & Further Reading

  • 2021 International Residential Code, R507.6, R507.5, and R507.9 with Tables R507.6, R507.5(1), and R507.9.1.3(1) — deck joist spans, beam spans, and ledger fastening
  • American Wood Council, DCA 6 — Prescriptive Residential Wood Deck Construction Guide (Tables 2 and 5)
  • 2021 IRC Table R401.4.1 — presumptive soil load-bearing values used for footing sizing