Window Replacement Cost Calculator

Insert or full-frame decides the price before anything else does

Every serious window estimator asks one question before asking anything about brands, glass packages or colours: is this an insert or a full-frame replacement? The answer moves the total by 40% to 60% on the identical window, and a calculator that never asks it is producing a number that cannot be checked against any real quote.

An insert — also sold as a pocket or retrofit replacement — takes out the sash, the balances and the stops, and slides a complete new window into the existing frame. The old frame, sill and exterior trim stay exactly where they are. Interior casing is untouched, siding is untouched, and a two-person crew can do several openings in a day.

A full-frame replacement removes everything down to the rough opening in the framing. That means the sill, the jambs, whatever flashing exists, and any rot found underneath. A new unit with an integral nail fin is set, flashed, foamed, and then interior casing and exterior trim are rebuilt on both sides. On a stucco, brick-veneer or fibre-cement wall this also means cutting and patching the cladding, which is why the exterior finish repair line in the calculator is not a rounding item.

The insert is cheaper for an honest reason and it charges for that in a currency people forget: glass area. Putting one frame inside another typically costs 50 to 75 mm across each dimension. On a large window that is a few percent of the view. On a 600 × 600 mm bathroom or stairwell window it is close to a fifth of the daylight, and the calculator reports that percentage for your specific list rather than leaving it as a footnote. Insert also assumes the existing frame is square, sound and dry. If it is not, an insert seals a rotting frame permanently inside the wall.

Material, labour, and the pile nobody puts in the brochure

The reason this tool separates material from labour is not tidiness. It is that the two behave completely differently. The price of a vinyl double-hung with a low-E argon package varies modestly between suppliers in the same city. The installed labour rate does not: two crews quoting the same street can be 40% apart, because one is a two-person owner-operated outfit and the other is a sales organisation with a call centre, a showroom and a commissioned rep who has to be paid out of your job.

On a typical residential replacement in the United States, labour lands between 30% and 50% of the total. Outside that band something is worth asking about. A quote where labour is only 20% has usually buried the installation inside the window price, and the result is a number you cannot compare with anyone else's. A quote where labour is 60% is either a builder-grade window being installed at premium rates, or a job with real complications that ought to be named on the estimate.

Then there is the third pile, which almost never appears in advertising and reliably appears on the final invoice: permit and inspection fees, removal and disposal of the old units, exterior finish repair, and access equipment. These are neither material nor labour in any useful sense, so the calculator gives them their own line.

Height is a labour multiplier and an equipment charge at the same time

Working at ground level is a stepladder and four paces to the truck. The second storey is an extension ladder, a harness, and roughly double the time to move each sash. The third storey and above is scaffold, a boom lift or a swing stage, and the crew is no longer setting windows — it is setting up.

Those are two separate costs and the distinction matters more than it sounds. The labour multiplier is per window: every opening upstairs takes longer than the same opening downstairs. The equipment is per job: scaffold rental, delivery and takedown cost the same whether you replace one upstairs window or seven. That single fact is the strongest financial argument for doing a whole floor in one go instead of one window a year. Spreading the job out means paying the setup charge over and over for the same wall.

Size does not scale with price, in either direction

People assume that if a 1.4 m² window costs $800, then a 2.8 m² window costs $1,600 and a 0.7 m² window costs $400. Neither end of that is true, and the reasons are different at each end.

Small windows cost more per square metre. Hardware, weatherstripping, balances, packaging, freight and the simple fact of the installer arriving at that opening are costs per opening, not per unit of area. That is why the calculator has a fixed price per opening as a separate editable field rather than folding it into the area rate: the moment you split it out, the reason a tiny bathroom window costs $480 becomes obvious.

Odd sizes cost more than round ones. Manufacturers build to a grid. A window within about 25 mm of a stock size comes off the shelf; a window 80 mm away from one goes into the made-to-order queue and carries a premium of roughly 10% to 15% — and, crucially, that premium is a step, not a ramp. Being 40 mm off a stock size costs almost as much as being 200 mm off. If you are ordering several windows and one dimension is close to a stock size, it is sometimes cheaper to adjust the rough opening than to pay for a custom unit on every one of them.

Oversize costs disproportionately more. Past roughly 1,800 mm wide or 1,500 mm tall, the unit leaves the standard manufacturing envelope and several things change at once: glass thickness has to increase for wind load, the frame needs reinforcement, hardware moves up a class, and the unit becomes a two-person lift. The cost grows faster than the area, which is why the tool applies a power-law premium rather than a flat percentage, and why the per-square-metre figure on your biggest window will be visibly higher than on the middling ones.

The practical consequence: dividing a quote's total by its total glass area to compare bids is meaningless. Two jobs with the same area can be legitimately thousands of dollars apart because of the mix of sizes and opening types. The calculator prints the cost per unit area on each individual line for exactly that reason — the three numbers will not match, and that is the lesson.

What double pane, low-E, argon and triple pane each actually buy

These are four separate upgrades that get sold as one word, and they do different things.

• Double pane cuts the centre-of-glass U-factor from about 1.04 to about 0.49 — the single biggest jump available, and more than half the total benefit of the whole stack. It does comparatively little for solar heat.
• Low-E coating takes the U-factor from about 0.49 to about 0.32, and takes the solar heat gain coefficient from about 0.61 to about 0.38. That second number is the important one in any climate with a real cooling season: low-E is doing more of its work against the sun than against conduction.
• Argon fill moves the U-factor from about 0.32 to about 0.28. Real, small, cheap, and it does nothing whatsoever for noise — a point worth remembering when it is sold as part of an "acoustic" package.
• Triple pane reaches about 0.18. It is the second doubling on a diminishing-returns curve: the same size of improvement in R-value terms buys roughly half what the previous step bought, while adding a third of the glass cost and a lot of weight. Its strongest argument in cold climates is not the energy bill — it is the temperature of the inner glass surface, which is what actually determines whether the room next to the window is comfortable and whether condensation forms.

One thing the glass cannot fix: the frame. An aluminium frame without a thermal break conducts heat around a thousand times better than uPVC, and the frame is 20% to 30% of the window's area. Putting a premium triple-glazed unit into a plain aluminium frame is spending on the wrong part of the assembly, and the calculator applies that penalty to the whole-window U-factor rather than quietly reporting the centre-of-glass number that appears on the sticker.

The energy payback, worked honestly

Here is the arithmetic the sales visit skips. Annual saving has two components, both derived rather than asserted. Conduction saving is area × ΔU × degree days × 24 ÷ 1000, giving thermal kilowatt-hours. Solar saving is area × ΔSHGC × annual irradiation × shading factor. Both are thermal energy, not purchased energy: they only become money after passing through the furnace efficiency or the air conditioner's COP, and skipping that division inflates the cooling saving by more than a factor of three.

Run ten single-pane windows in aluminium frames up to double-pane low-E argon in vinyl, in a climate with about 4,700 heating degree days, and the honest answer is around $400 a year. Against an insert job that is roughly a 20-year payback. Against a full-frame job it is closer to 30 — which is longer than any warranty on the product and close to the service life of the window itself. Over the whole 25-year service life the energy saving typically returns somewhere between 70% and 95% of what the job cost. In plain terms: it does not pay for itself.

That is not a reason to keep leaky single-pane windows. It is a reason to buy windows for the right reasons and to treat the energy number as a contribution rather than a justification. Comfort next to the glass, condensation and the mould that follows it, noise, security, operation that does not need a shoulder, and the fact that a rotten sill is a structural problem — those are what actually make a replacement worth doing. If the only argument in the brochure is the energy bill, and the brochure quotes a payback under ten years, ask which degree-day figure and which equipment efficiency produced it.

There is one genuine saving this calculation deliberately does not estimate: air leakage. A warped 60-year-old sash leaks, and a new unit does not. That gain is real and can be substantial, but quantifying it honestly needs a blower-door test rather than an assumption, and inventing a number here would be exactly the behaviour this page criticises.

How to read three quotes that all say the same total

Once material and labour are separated, comparing bids stops being guesswork. Work through them in this order.

• Insert or full-frame? If two quotes differ on this, they are not competing offers and the cheaper one is not cheaper. Get both priced the same way.
• What exactly is the glass package? Ask for the NFRC label numbers: U-factor, SHGC and visible transmittance, for the whole window and not the centre of glass. "Energy efficient" is not a specification.
• What share is labour? Outside 30%–50%, ask why. The answer is usually legitimate and always informative.
• Who pays for rot? Full-frame quotes should state a per-opening allowance for replacing damaged framing, because nobody knows what is behind the trim until it comes off. A quote silent on this will produce a change order.
• Disposal, permit and finish repair. Named separately, or absent? Absent does not mean free.
• Warranty terms. Glass seal failure, hardware, frame and labour usually carry four different periods. A twenty-year glass warranty with a one-year labour warranty means the seal is covered and the trip to fix it is not.

Everything is computed in your browser

No upload, no account, nothing logged. Every price in the tool is an editable field pre-filled with an order-of-magnitude starting point, not a hidden constant, so a contractor can paste in real numbers and get a real estimate. The arithmetic lives in a separate module that is verified outside the browser against hand-worked examples — including the insert-versus-full-frame percentage difference, the labour share band, the non-linear size premium and the degree-day energy saving above — so the figures are checkable rather than merely confident.