Stair Calculator

The riser you type is a wish. The riser you get is a division.

Every stair starts from one number that cannot be negotiated: the vertical distance between the floor you leave and the floor you arrive at. Everything else bends around it.

You cannot build 15.43 risers. So the tool divides the total rise by your target riser, rounds to the nearest whole number, and then divides the total rise by that whole number again. A 108 in rise with a 7 in target gives 15 risers of 7.2 in. Ask for 7.5 in and you get 14 risers of 7.71 in — still legal under the IRC, but noticeably steeper, and the tread has to shrink to keep the stride sensible.

That second division is the step most calculators skip. They show you the riser you asked for and quietly leave the remainder somewhere. In a real stair the remainder has nowhere to hide: it ends up in one step, usually the top or the bottom one, and that step is now different from all the others.

Uneven risers are not a cosmetic problem. Walking a flight is a rhythm — after two steps your body stops looking and starts predicting, and it predicts the spacing it has already felt. A single riser that is half an inch off breaks the prediction at the exact moment you are least prepared for it. This is why codes cap the variation between the tallest and shortest riser in a flight at 3/8 in (about 9.5 mm), and why every riser this tool reports is identical by construction rather than by luck.

There is always one more riser than there are treads

A stair that lands on the upper floor has its last riser topped by the floor, not by a tread. The floor is the arrival surface. So the count is:

risers = treads + 1

Fifteen risers, fourteen treads. The total run is fourteen tread depths, not fifteen. At 10.5 in per tread that is a 12 in difference — a foot of floor that either exists or does not, and the kind of error that only shows up when the last tread lands past the wall.

The exception is a stair that ends at a constructed landing or platform where the top step is a real step you build. Then treads equal risers. The tool has a switch for it because both cases are common and mixing them up costs exactly one tread depth in either direction.

Blondel: the check almost nobody shows you

In 1675 François Blondel noticed that climbing costs about twice as much effort per unit of height as walking costs per unit of distance, and that a comfortable stride on the flat runs about 25 in. Put those together and you get the rule that has survived three and a half centuries:

2 × riser + tread ≈ 24 to 25 in (63 to 65 cm)

A 7 in riser with a 10.5 in tread gives 24.5 in — comfortable. A 7.75 in riser with a 10 in tread gives 25.5 in, which is legal in US residential work and still feels like you are reaching. A 6 in riser with a 14 in tread gives 26 in and produces the other failure mode: the stair where your foot keeps landing early and the flight feels endless.

Two things follow from the formula that are not obvious. First, riser and tread are not independent choices — pick one and the other is nearly determined. Second, a stair can be fully code-compliant and still be an uncomfortable stair, because codes set maxima and minima while Blondel describes a relationship. That is why this tool reports the Blondel result as its own line rather than burying it.

A useful second opinion is the sum rule: riser plus tread between 17 and 18 in. When both rules agree, the stair is good. When they disagree, something in your inputs is at an extreme.

The finish flooring moves the whole stair

Framing happens before flooring. So the height you measure on site is subfloor to subfloor, and the height the stair will actually have to climb is subfloor to subfloor plus whatever goes down upstairs, minus whatever goes down downstairs.

Three quarters of an inch of hardwood upstairs and nothing downstairs means the finished rise is 3/4 in taller than the number you framed to. Tile with a mortar bed can be more than an inch. That difference does not distribute itself across fifteen risers — it lands entirely on the top one, which becomes 3/4 in taller than its fourteen identical siblings. You are now past the 3/8 in tolerance, on a step positioned exactly where people stop concentrating because they can see the floor.

The mirror image happens at the bottom. Flooring laid on the lower level raises the point where the first riser starts, shortening it by the same amount. Both fields are in the tool for that reason, and the tool flags the situation when the thickness exceeds the tolerance.

Short opening, steep stair — and the point where it stops working

The most common thing that goes wrong with a stair is not arithmetic. It is that the opening was framed first, and the stair had to fit inside it.

Squeezing a stair into a short run means fewer treads, which means fewer risers, which means each riser gets taller. That works until it does not, and there are two hard limits closing in from opposite directions:

• The run sets a maximum riser count. Each tread has to stay at or above the minimum depth — 10 in in US residential work, 11 in commercial. Divide the available run by that minimum and you have the largest number of treads that fits, and therefore the largest number of risers.
• The rise sets a minimum riser count. Each riser has to stay at or below the maximum height — 7 3/4 in residential, 7 in commercial. Divide the total rise by that maximum, round up, and you have the smallest riser count that is legal.

When the maximum falls below the minimum, no straight stair exists in that footprint. No amount of adjusting will produce one, and a calculator that answers anyway is lying to you. This tool says so and prints both numbers, because knowing that the run allows 11 risers while the height demands 16 tells you exactly how far off you are — and whether the fix is a longer opening, an L-shaped flight with a landing, or winders.

Headroom is the dimension you find out about with your forehead

Headroom is measured vertically from the line joining the tread nosings to whatever is above — the ceiling, or the edge of the opening in the upper floor. Six feet eight inches is the residential minimum. It is not measured from the treads to the floor above; it is measured along the whole climb, and that is what makes it interesting.

The stair rises. The underside of the upper floor does not. So the clearance shrinks by one riser height with every step, and at some point on the flight it drops below the minimum. Everything past that point has to be under open air, which means the opening in the upper floor has to start there.

The arithmetic is simple and almost never done in advance: take the height of the floor underside, subtract the minimum headroom, divide by the riser height, and round down. That is how many steps can pass beneath the floor. The rest of the run is the length of opening you need to frame. Deeper joists eat into it. A thick assembly with a shallow stair can push the opening several feet longer than people expect, and discovering that after the joists are cut is an expensive afternoon.

Landings, width, handrails

Three constraints sit outside the riser-and-tread arithmetic but decide whether the stair is buildable at all.

Above the maximum rise per flight a landing becomes mandatory, and a landing is not free: it consumes at least its own depth of run, usually the full width of the stair. A stair that was already tight for space becomes an L or a U at that point, not a longer straight run.

Handrail height is measured from the nosing line, not from the treads, which is why it sits lower than it looks on a drawing. Width is measured between the finished surfaces — above the handrail, between the walls; at handrail height, from the rail face — so a rail that projects into the flight subtracts from the clear width you are allowed to count.

Metric and imperial, and the conversion trap

The tool works in either. Switching units converts what you already typed rather than reinterpreting it, which sounds obvious and is where a lot of tools quietly fail: reinterpreting turns a 280 into 280 inches and the stair explodes with nothing on screen explaining why.

Internally every calculation runs in millimetres, converting once on the way in and once on the way out. Mixing units inside the arithmetic is the classic source of construction-tool bugs, because the wrong answer still looks like a plausible answer.

The rule sets differ in more than notation. Brazil's NBR standards put risers in a 16 to 18 cm band with treads from 28 to 32 cm — narrower on both ends than the IRC, which sets only a maximum riser and a minimum tread and lets everything else float. A stair that is routine in one country can be non-compliant in the other while being physically identical, so pick the rule set you are actually building under.

How to measure the rise so the rest is right

Measure the vertical distance at the point where the stair will actually sit, not at the nearest convenient wall. Floors are not level and slabs are not flat; a 10 mm difference across the room becomes a 10 mm error on the last riser.

Take the measurement from the surface people will stand on to the surface they will arrive at. If either finish is not down yet, measure the substrate and put the thicknesses into the two adjustment fields rather than doing the sum in your head — the whole reason those fields exist is that the correction is easy to remember at the desk and easy to forget on site.

Then check the drawing against the room. The profile view shows the total run against the space you entered, the stringer as the diagonal you will actually have to cut, and the pitch angle. Most stairs that turn out badly were dimensionally correct and geometrically obvious — someone just never looked at them from the side.

Privacy

Everything runs in your browser as arithmetic. Nothing is uploaded, nothing is logged, and there is no account. Close the tab and the numbers are gone.