Framing Takeoff Guide

A framing takeoff turns a plan set into a lumber order: a count of studs by length, linear feet of plate stock, a header list by opening, joists and rafters by size and span, sheets of panel goods, and a pile of hardware. The fastest honest starting point for walls is the rule every framer knows: one stud per linear foot of wall at 16 inches on center. That is not the spacing math, which gives 0.75 studs per foot. The extra quarter stud per foot is the corners, the wall intersections, the king and jack studs at openings, and the cripples. On a normal house with normal rooms, those extras land close enough to a quarter stud per foot that the rule holds.

Framers on ContractorTalk describe the whole method in about one sentence: take a stud per foot of wall, multiply joist and rafter runs by 0.75 and add one, figure sheathing at 32 square feet a sheet, then add up headers and nails. That is a real production takeoff, and it is fast because framing lumber is a commodity where being 5 percent long costs very little and being 5 percent short costs a crew standing around. The detail work is not in the studs. It is everywhere else.

The Six Counts

A framing takeoff is six lists that do not share units, and they should stay on six separate lines:

• -Studs as a count, broken out by length and by size (2x4 versus 2x6)
• -Plate stock in linear feet, converted to sticks of a buyable length
• -Headers and opening framing as a per-opening schedule, not a bulk number
• -Horizontal framing (joists, rafters, beams, posts) as a count by size and length
• -Panel goods as sheets, by thickness and span rating, split by wall, roof, and floor
• -Hardware and fasteners as counts: hangers, ties, straps, hold-downs, nails by box

Roll any two of those together and you lose the ability to price an odd job. A tall-wall job and a standard job can have identical square footage and completely different stud counts, because stud length drives cost and stud count does not move.

Work in Build Order

Take the frame off in the order it goes up: floor system, walls by level, then roof. This does two things. It groups your quantities by the load that will actually be delivered, so the takeoff doubles as a delivery schedule. And it forces you to think about what is holding up the thing you just counted, which is where the posts, beams, and hangers live. Estimators who take off by sheet instead of by sequence reliably find the walls and miss the point loads under them.

Know Who Is Doing This Before You Start

One thing worth settling early: many framers do not do their own lumber takeoff at all. The recurring advice on contractor forums is to let the lumber yard run the material list, because the yard does it free, the yard carries the liability if the list is short, and the yard is more likely to waive a restock fee on returns from a list it generated. That is a legitimate strategy, not a shortcut. But it only works if you can check their list, which means you still need the math below. A yard list you cannot sanity-check is a number you are signing for blind.

Wall framing is the biggest count and the easiest to get approximately right. Start with field studs, then add the framing that spacing does not account for.

Field Studs by Spacing

The base count for a straight run of wall is the wall length divided by the spacing, plus one for the stud at the far end:

Field studs = Wall length (in) / Spacing (in) + 1

Per linear foot, that reduces to a constant you can multiply straight through a wall schedule:

Note 19.2 inches on center. It is five layouts per 8 foot sheet and it shows up on engineered floor systems and on advanced framing plans. If a plan calls it out and your takeoff assumes 16, you will order roughly 20 percent more studs than the job needs.

Then Add the Framing Spacing Misses

Corners, intersections, and openings all take studs the spacing math never sees. The common add-ons, which match what Autodesk publishes in its lumber takeoff walkthrough and what Buildxact uses:

Two notes on that table. A 90 degree corner takes three studs in a conventional California corner or three-stud corner, but advanced framing details use two studs plus drywall clips and save the third. If the plan shows an advanced framing detail, count what the plan shows. And cripples are the count people skip because each one is short. On a wall with eight windows they are not a rounding error.

Worked Example: One Exterior Wall Run

A 48 ft exterior wall, 2x6 at 16 inches on center, 9 ft plate height, with two 90 degree corners, one interior wall teeing in, and three 3 ft wide windows.

• -Field studs: 48 x 0.75 = 36, plus 1 = 37
• -Corners: 2 x 3 = 6
• -Intersection: 1 x 2 = 2
• -Openings: 3 windows x 2 sides x 2 studs = 12
• -Cripples: each 3 ft opening is 36 in wide, so 36 / 16 = 2.25, call it 2 above and 2 below, times 3 windows = 12
• -Total: 69 studs at 10 ft (9 ft plate height plus plates needs a 10 ft stud, or a precut)

Compare that to the one-stud-per-foot shortcut: 48 studs. The detailed count came in 44 percent higher, because this wall is opening-heavy and corner-heavy. That is the honest limit of the rule of thumb. One stud per foot is calibrated to a whole house where long blank walls average out the busy ones. Applied to a single elevation full of glass, it is badly short. Use the shortcut for a whole-structure ballpark and the detailed count when you are pricing a specific wall.

Plate Stock

Standard load-bearing wall framing is a single bottom plate and a double top plate, which is three plate lines running the full length of every wall:

Plate LF = Total wall LF x 3

Multiply by 2 instead where the plan shows a single top plate, and by 4 where you have a double bottom plate on a stepped foundation or a pressure-treated sill plus a bottom plate. Then convert to sticks. Plate stock is bought long, usually 16 ft, to keep splices down:

• -340 LF of wall x 3 = 1,020 LF of plate
• -1,020 / 16 = 63.75, round to 64 sticks of 2x 16 ft
• -Pressure-treated bottom plate where it sits on concrete: that is a separate line item at 340 LF, or 22 sticks, and it is a different price than the rest

That last point is a classic miss. The bottom plate on a slab or a foundation wall has to be pressure treated or naturally durable, and plenty of takeoffs bury it in the total plate footage at untreated pricing. Break it out.

Stud Length Is a Line Item, Not a Detail

Precut studs exist because the arithmetic is annoying. A 92 5/8 in precut plus a 1.5 in bottom plate plus two 1.5 in top plates gives a finished 8 ft 1 1/8 in wall, which lands a full 8 ft sheet of drywall with clearance. The 104 5/8 in precut does the same for a 9 ft wall. If you are ordering 8 footers and cutting them, you are paying for waste and labor to get to the same place. Check the plate height on the plan, pick the precut, and put the odd walls on a separate line.

Openings are where framing takeoffs go wrong, because an opening is not one item. It is a header, two king studs, two or more jack studs, cripples above, a rough sill and cripples below on a window, and sometimes hardware. Count them as an assembly per opening.

You Cannot Size a Header from the Opening Width Alone

This is the single most common mistake in framing estimating. Per IRC Section R602.7, header size comes out of Table R602.7, which is indexed on the clear span of the opening, the width of the building, and the ground snow load, and is split by whether the header carries roof only, roof plus one floor, or floor only. The table runs building widths of 20, 24, 28, 32, 36, and 40 ft and ground snow loads of 30, 50, and 70 psf. Same 6 ft window, different building width, different header.

The 2024 IRC edition clarified something worth knowing because it changes numbers: the building width in the table is the overall dimension from exterior wall face to exterior wall face in the direction of the span, not the distance between bearing walls. Practitioners who used the shorter number were understating the load and undersizing headers.

What that means for a takeoff: if the plan has a header schedule, take off the header schedule. Do not derive headers from a span table yourself unless you are certain of the load path, and never price a job off a header size you guessed from the window width.

Header Stock Math

For a conventional built-up header, the buyable length is the opening width plus the bearing on each end. The common shortcut, which shows up in both the Autodesk and Buildxact walkthroughs, is opening width plus 7 inches, which covers 1.5 in of jack stud bearing on each side plus tolerance. IRC R602.7 requires a minimum of 1.5 inches of bearing at each end onto the jack studs.

A built-up 2x header in a 2x6 wall is two members plus a spacer to make up the 5.5 in wall thickness. That spacer is usually 1/2 in plywood ripped to header depth, and it is real material that belongs on the list. In a 2x4 wall, two 2x members plus a 1/2 in spacer comes to 3.5 in and no extra is needed.

The right-hand column is the one that matters on the order. You do not buy a 6 ft 7 in header. You buy an 8 footer and eat the drop, or you get two 6 ft 7 in headers out of a 14 ft stick if you plan the cut. That is a cut list decision, and it is where a framing takeoff turns into money.

When the Header Becomes Engineered Lumber

IRC permits LVL as an alternative to sawn headers, and LVL is usually shallower for the same span. That matters above a wide opening where a deep sawn header will not fit under the plate. Wide openings, garage door headers, and anything carrying a point load from above tend to end up as LVL, PSL, or a steel flitch. These are priced per linear foot by depth and ply count, they are ordered rather than stocked, and lead time is a scheduling item. Put engineered headers on their own line with the ply count, because a 3-ply 14 in LVL and a 2-ply 11 7/8 in LVL are not remotely the same buy.

The Per-Opening Checklist

For every door and window on the plan, confirm you have counted:

• -Header, at the size the schedule calls out, in the stock length you will actually buy
• -Header spacer if it is a 2x6 wall
• -Two king studs, full height
• -Jack studs each side, more than one where the header schedule or R602.7.1 requires it
• -Cripples over the header, at wall spacing
• -Rough sill on windows, and cripples under it at wall spacing
• -Hangers or straps if the header is hung rather than bearing
• -For wide garage openings, the hold-downs and the portal frame sheathing and strapping if the plan uses one

Miss the portal frame detail on a narrow garage-front wall and you have missed sheathing, straps, hold-downs, and anchor bolts on the one wall where the inspector will definitely look.

Horizontal framing counts the same way walls do, with one difference: the members are long, so length matters more than count, and the accessories are a bigger share of the price.

Floor Joists

Joist count for a bay is the span perpendicular to the joists divided by spacing, plus one:

Joists = Bay width (ft) x 12 / Spacing (in) + 1

At 16 inches on center that is the same 0.75 per foot constant as studs, which is why framers say "times point seven five, round up, add one" for joists and rafters both. Then add the pieces that are not field joists:

• -Doubled joists under every parallel partition wall above, and around every opening
• -Headers and trimmers at stair openings, chases, and hearths
• -Rim or band joist around the full perimeter, in linear feet
• -Blocking or bridging at mid-span where the plan or code requires it, typically a row per bay
• -Squash blocks and rim closures if the floor is an I-joist system

That last one catches people moving from sawn lumber to I-joists for the first time. An I-joist floor is not just joists. It is joists, rim board, web stiffeners, squash blocks at point loads, and hangers, and the manufacturer's installation guide specifies which are required where. Weyerhaeuser publishes a Trus Joist floor framing standards guide for exactly this, and the accessories in it are not optional.

Rafters and Roofs

Rafter count works like joists, off the run of the roof. Rafter length is where the slope comes in:

Rafter length = Horizontal run x Pitch factor

The pitch factor is the hypotenuse of the slope triangle, the square root of 1 plus the rise over 12 squared. It applies to rafter length and it applies to roof area for sheathing:

Add the overhang to the run before you multiply, and remember the rafter needs to be bought in a stock length that covers the full sloped length plus the plumb cut and tail. A 15 ft 8 in rafter is a 16 ft stick, and a 16 ft 10 in rafter is an 18 ft stick with almost 14 inches of drop per rafter. That drop across forty rafters is real lumber, and it is why rafter length should drive which stock length you buy rather than being rounded up quietly in a spreadsheet.

Also count, because these are the roof items that vanish:

• -Ridge board or ridge beam, and they are not the same thing or the same price
• -Hip and valley rafters, which are longer than commons and usually one size deeper
• -Jack rafters at hips and valleys, each a different length
• -Collar ties and rafter ties at the spacing the plan calls out
• -Ceiling joists, which are a separate count from rafters
• -Sub-fascia, barge rafters, lookouts, and soffit framing around the full roof edge

Trussed Roofs

If the roof is trussed, the truss package is a supplier quote off the truss plan and not something you take off stick by stick. What you do still take off is everything the truss company does not ship: the lateral and permanent bracing the truss design drawings require, the gable end bracing, the blocking between trusses at the bearing wall, the sub-fascia, and the hardware. One framer's working checklist posted on the JLC forums handles it this way: the truss plan determines the 2x4 bracing footage, roof edge footage determines rough fascia and soffit block, and truss count times overhang length determines soffit joist footage. That is the right way to think about it. The trusses are a line item you get quoted; the framing around them is yours to count.

Beams and Posts

Every point load lands somewhere. Trace the beams and posts down from the roof to the foundation and make sure each one is on the list, at the ply count the plan shows, along with its column caps, bases, and hold-downs. A missed 4-ply LVL beam is a four figure hole in a bid, and it is missed because it lives on the structural sheet and the estimator was working off the floor plan.

Panel goods are the easiest arithmetic in the whole takeoff and the place where the wrong product gets ordered most often.

The Count

A standard panel is 4 ft by 8 ft, which is 32 square feet:

Sheets = Net area (SF) / 32, rounded up

Net area means after deducting openings, and whether you deduct at all is a judgment call. Buildxact works an example that deducts: a 20 ft by 8 ft wall is 160 SF, minus a 4 ft by 5 ft window at 20 SF, leaves 140 SF, divided by 32 is 4.4, so five sheets. That is correct on paper. In the field, crews routinely sheathe straight over window openings and cut them out after, because it is faster and it keeps the shear panel continuous while the wall is standing. If that is how your crew works, do not deduct small openings at all. Deduct garage doors and large glass walls, which are too big to sheathe over.

The honest version: deduct openings larger than about 32 square feet, ignore the rest, and let them serve as your waste factor on the walls.

Roof Sheathing Uses Sloped Area

Roof panel count comes off the sloped area, not the footprint. Take the plan area, multiply by the pitch factor from the table above, then divide by 32. A 1,800 SF roof footprint at 6/12 is 2,012 SF of actual roof, which is 63 sheets before waste instead of the 57 you would get off the flat number. That is a six sheet miss on a modest house and it grows with pitch.

Get the Span Rating Right

This is the part that costs money when it is wrong. APA Rated Sheathing carries two numbers, and estimators conflate them constantly. The Performance Category is the thickness callout. The Span Rating is the two numbers separated by a slash, where the first is the maximum roof support spacing and the second is the maximum floor support spacing, both in inches. Per APA's published datasheet:

So a plan calling for trusses at 24 inches on center and a 7/16 panel is fine for the roof. The same 7/16 panel is not a subfloor panel at 24 inch joist spacing. Order by the span rating the plan requires, and put wall, roof, and floor panels on three separate lines even when the thickness happens to match, because they usually do not.

Two more product distinctions worth holding onto:

• -Structural I is a higher-grade rated sheathing with better racking and cross-panel strength, specified for engineered shear walls and diaphragms. If the shear wall schedule says Structural I, a standard rated panel is not a substitute, and APA notes it may not be locally available. That is a lead time problem, not just a price problem.
• -CDX is not exterior plywood. APA flags this explicitly: the Exposure 1 bond classification that the trade calls CDX is regularly mistaken for an Exterior bond and used in applications it does not qualify for. Exposure 1 tolerates construction delays before the building is dried in. It is not a long-term weather-exposed product.

Subfloor and the Glue

Subfloor is usually tongue and groove, which changes the coverage slightly and changes the handling a lot. Count sheets the same way, then add the items that ride with them: construction adhesive by the tube, subfloor screws or ring shank nails by the box, and any Sturd-I-Floor upgrade the plan specifies. Adhesive gets skipped in takeoffs at a remarkable rate for something that is required by nearly every subfloor installation spec.

This is the part of the job a takeoff tool earns its keep on. Pulling wall areas, roof areas, and opening deductions off a plan set and converting them to sheet counts is mechanical work, and Tectonic does that side of it, reading quantities straight off the PDF and applying your overhead and profit markup to the material total. Do a set by hand first anyway. If you have never counted panels off an elevation yourself, you will not notice when a tool sheathes a wall that was supposed to be a shear panel of a different grade.

Framing runs on four units and everyone mixes them up at least once. This section is the reference to keep open while you work.

Nominal Versus Actual

Nominal is the size the board was cut to green, before drying and planing. Actual is what shows up on the truck. Per Lowe's dimensional lumber reference, a nominal 2x4 measures 1 1/2 in by 3 1/2 in. You order and price nominal; you build and lay out actual.

Notice the width loss goes from 1/2 in on a 2x4 to 3/4 in on a 2x8 and wider. That is why a 2x8 rim does not sit flush with a 2x6 wall the way people expect, and it is why layout dimensions taken off nominal sizes drift.

Board Feet

A board foot is 1 in thick by 12 in wide by 12 in long. The formula uses nominal dimensions:

Board feet = Nominal thickness (in) x Nominal width (in) x Length (ft) / 12

• -A 2x4x8 is 2 x 4 x 8 / 12 = 5.33 bd ft
• -A 2x6x10 is 2 x 6 x 10 / 12 = 10 bd ft
• -A 2x10x16 is 2 x 10 x 16 / 12 = 26.67 bd ft

You mostly do not need board feet for residential framing, because dimensional lumber is priced by the piece or by the linear foot. Board feet matter when you are buying in volume, comparing a price to a published lumber market number, or working with timbers and hardwoods. Know the formula, then use piece counts.

The Units Table

Stock Lengths and the Cut List

Dimensional lumber comes in even lengths, usually 8, 10, 12, 14, 16, and 20 ft. Your takeoff should pick the stock length that cuts cleanly into what you need, and this is where a takeoff stops being arithmetic and starts saving money. Autodesk's lumber takeoff walkthrough makes the point well: if you need 6 ft members, buy 12 footers and cut them in half rather than buying 8 footers and throwing away 2 ft on every piece. Same idea for 7 ft walls out of 14 footers.

Run the check on every repeated member:

• -What length do I need, including cuts and bearing?
• -What stock lengths divide into that with the least drop?
• -Is the drop long enough to become blocking, or is it trash?

That last question is the one that separates a framing takeoff from a materials list. On a well-planned job the drops become blocking, backing, and cripples, and the blocking line on the order gets smaller. On a badly planned job the drops become a dumpster line item.

Here is the honest state of framing waste factors: the published numbers do not agree, and the disagreement is not small.

Nobody Agrees on the Number

Takeoff walkthroughs from software vendors commonly apply 15 percent to stud counts and 5 to 10 percent to plates. Waste factor reference tables aimed at estimators put framing much lower, with EZ-Estimates listing wall framing at 5 to 8 percent, floor joists at 3 to 5 percent, roof framing at 5 to 8 percent, and sheathing at 5 to 8 percent.

And then there are the builders. On the JLC forums, a contractor laying out his framing takeoff method finished with "add 15 to 20 percent to all totals," noting his yard would take returns back without a pickup or restock fee. A forum moderator came back immediately: at that level of takeoff detail, 15 to 20 percent is "awfully excessive," and the extra lumber gets stolen, or warps in the sun, or goes back to the yard anyway. A custom home builder in Texas added that 20 percent waste is excessive on some items and that his yard would not take material back unless it was in excellent condition, and even then there was a hefty restock fee.

They are all right, because the correct waste factor is not a property of the material. It is a property of your return policy and your crew. If your yard takes clean returns free, ordering long is nearly risk free and 15 percent is rational. If your yard charges a restock fee and your lumber sits in the weather, 15 percent is money set on fire.

Use This Instead of a Single Number

Two of those rows matter more than the rest. Trusses get zero waste, because they are manufactured to a shop drawing and there is no such thing as a truss offcut. Applying a blanket 10 percent across a whole framing package quietly pads the truss line, which is often the single biggest number in the package. And a cut roof is genuinely wasteful in a way that walls are not, because hip, valley, and jack rafters produce drops in lengths that fit nothing else. That is why the same house framed with trusses and framed stick-built do not carry the same waste.

The underlying economics, which the OrbitalJump waste reference states plainly, is that running short is usually more expensive than ordering long. A crew standing around waiting on a supply run costs more in an hour than the lumber you saved. That asymmetry is the reason waste factors exist. It is not a reason to stop measuring.

What Actually Gets Missed

Missed studs are rare. Here is what actually falls off framing takeoffs, roughly in order of how often it happens:

• -Blocking and backing. Fire blocking in tall walls and at floor lines, blocking for cabinets, grab bars, closet rods, handrails, TV mounts, and stair skirts. It is all short pieces of 2x, it is all on the plan somewhere, and it is almost never on the order.
• -Temporary bracing. One framer's list on JLC budgets 20 to 30 2x4x16s for temp bracing and notes they turn into blocking later. That is the right instinct: the material is real, and it is reusable, so it should be on the list once rather than twice or never.
• -Hardware. Joist hangers, hurricane and seismic ties, straps, hold-downs, column caps and bases, anchor bolts. On a plan with a connector schedule, the schedule tells you exactly what to count, and connector schedules are often on a structural sheet an estimator working off architecturals never opens. In high wind and seismic zones this line is not small.
• -Fasteners. Framing nails, sheathing nails, hanger nails (which are a specific product, not a substitute), subfloor adhesive, subfloor screws. Nails are cheap and hanger nails are not optional, since connectors only reach their rated capacity with the specified fastener in the specified holes.
• -The pressure-treated line. Sill plates and any bottom plate on concrete. Different product, different price, regularly buried in the untreated plate total.
• -Rake walls and tall walls. A gable end wall with a sloped top plate is not a rectangle and its studs are all different lengths. Spreadsheets that take wall length times a per-foot constant get the count roughly right and the lengths completely wrong.
• -Ceiling strapping and soffit drops. The same JLC checklist treats every dropped ceiling soffit as a three-plate wall and counts it accordingly, and figures ceiling strapping at roughly 0.8 times the ceiling area in linear feet. Dropped soffits in kitchens and halls add up fast and appear on the reflected ceiling plan, which is another sheet estimators skip.
• -Stairs. Stringers, treads, risers, and the framing around the stair opening. Stairs are a small area on the plan and a disproportionate amount of material and cutting.

Close the Loop

The best framing estimators are not the ones with the best waste percentages. They are the ones who write down what they ordered, get the actual from the field, and compare. Track ordered versus installed by category across a season and your waste factors stop being borrowed from a blog and start being yours. That is the only version of this number that is worth anything, and it costs nothing but the discipline to record it.

Digital takeoff handles the mechanical half of this work, which is measuring walls, deducting openings, and converting areas to counts. Tectonic reads material quantities off a plan set and applies your markup for overhead and profit, so the hours go into the judgment instead of the arithmetic: which waste factor fits this crew, whether the connector schedule is fully counted, and whether that gable wall is really a rectangle. The counting was never the hard part of a framing takeoff. Noticing what the plan did not put in front of you is.