Two carpenters cut crown for the same 90-degree inside corner. One sets the saw to 31.6 degrees miter and 33.9 degrees bevel. The other sets it to 35.3 and 30.0. Both are right. The first piece is 38-degree crown; the second is 45-degree. Same corner, same room, settings that are off from each other by almost four degrees on the miter and four on the bevel. If you grabbed the wrong number off a chart because you didn't notice which crown profile it was written for, the joint opens up and you find out at the wall.
That gap is the part of crown molding that the printed cheat-charts never explained well. A corner angle alone doesn't tell you where to set the saw. The profile of the molding is half the equation, and the two halves interact in a way that isn't intuitive until you've seen the geometry behind it.
The corner is only one of the two inputs
A crown molding joint at a wall corner is a compound-angle cut. That word "compound" is doing real work. You're making two adjustments to the saw at the same time, and neither one can be set without the other.
The first is the miter: how far the blade swings left or right across the table. The second is the bevel: how far the blade tilts off vertical. A plain baseboard or a flat picture frame is a single-angle cut, just a miter, because the stock lies flat on the saw table. Crown doesn't lie flat. It springs out from the wall at an angle, so to cut it you have to account for both the corner you're wrapping and the tilt the molding already sits at.
The molding's tilt is the spring angle, and it's the input most people forget is an input at all. Standard crown comes in a few spring angles: 38, 45, and 52 degrees are the common ones, with 38 being the most widely stocked in North American big-box stores. Those numbers describe how steeply the crown leans out from the wall. A 38-degree crown lies back closer to the wall; a 52-degree crown stands up more toward the ceiling. The steeper the spring, the more bevel and less miter the same corner needs, because more of the corner's turn is being absorbed by the tilt of the blade instead of the swing.
What the trig is actually doing
The closed-form formulas for an inside corner are short:
- Miter angle:
M = arctan(sin φ × tan(C/2)) - Bevel angle:
B = arcsin(cos φ × cos(C/2))
Here φ is the spring angle and C is the measured corner. You never have to touch these to cut crown, but reading them once explains why the chart values move the way they do.
Notice the C/2. The saw only ever cuts half the corner, because the matching piece on the other wall cuts the other half. So a 90-degree corner feeds 45 degrees into both formulas, not 90. The spring angle φ then shows up as sin φ in the miter and cos φ in the bevel. Sine and cosine pull in opposite directions as the angle grows. That's the mathematical reason a steeper crown trades miter for bevel: as φ climbs, sin φ grows and cos φ shrinks, so the miter rises and the bevel falls. The two settings are linked through the same profile angle, which is why you can never look up one without knowing the other.
Plug in 38 degrees of spring and a 90-degree corner and you get 31.6 and 33.9. Plug in 45 degrees and you get 35.3 and 30.0. The corner never changed. The profile did, and the profile is what moved both numbers.
The convention that wrecks more cuts than bad math
Here is the trap that no amount of arithmetic will catch. There are two ways people state a spring angle, and they don't agree.
The trade convention measures the spring angle from the wall: lay a bevel gauge flat against the wall and read the angle up to the face of the crown. By that convention a common crown is 38 degrees. But some references, some saw fences, and some tools measure the same molding from the ceiling, and that flips the number to its complement. A 38-degree-from-the-wall crown reads as 52 from the ceiling. A 45 reads as 45 either way, which is exactly why 45 hides the problem until you switch to a different profile and the two conventions suddenly diverge.
| Crown profile | Spring from the wall | Spring from the ceiling |
|---|---|---|
| Common 38° crown | 38° | 52° |
| Steep 52° crown | 52° | 38° |
| Symmetric 45° crown | 45° | 45° |
If you enter the ceiling number into a formula that expects the wall number, every cut in the room is wrong by a consistent, confusing amount. The joints don't fail in an obvious way. They sit slightly proud or slightly gapped, and you chase the error around the room blaming your corner measurements when the real problem was a single mislabeled input at the start. This is the single most documented source of crown-molding error, and it's a vocabulary problem, not a math problem.
Real corners aren't 90 degrees, and the charts assume they are
Walk into any house older than a few years and put an angle finder in the corners. You'll read 88, 91, 89.5, 92. Drywall mud builds up in corners, framing settles, and almost nothing comes out at a clean 90. The Joe Fusco printed charts that carpenters carried for two decades, and most of the static tables online, list values only for 90-degree corners. They have no row for 91.
A two-degree error in the corner doesn't sound like much until you remember the saw only cuts half of it, and then the molding spans two walls, so the small error shows up doubled at the joint. On 45-degree crown, moving the corner from 90 to 88 shifts the miter from about 35.3 to roughly 34.5 and the bevel a few tenths the other way. That's the difference between a joint you caulk once and a joint you recut. The only way to land it is to measure the actual corner with a digital angle finder and run the real number through the formula, instead of rounding to the nearest chart entry.
Inside and outside aren't the same cut mirrored
It's tempting to think an outside corner is just an inside corner flipped. The angle math is related, but the saw setup is genuinely different, and treating them as interchangeable wastes stock.
For an outside corner the corner's effective angle is the reflex side, so the formula substitutes (360 − C) for the corner, or equivalently works from the supplement. More importantly, the molding gets oriented differently against the fence and the keeper piece is on the opposite side of the blade. An inside corner keeps the long point of the molding at the back; an outside corner keeps it at the front. Cut an outside corner with the inside-corner orientation and the angles might even be close, but the bevel leans the wrong way and the profile won't wrap the corner.
The practical consequence: running crown around a single room means alternating between two distinct setups, sometimes within a few feet of each other, and each one has its own miter, its own bevel, and its own way of holding the piece on the saw.
Knowing the numbers still isn't enough
This is the part that separates a usable answer from a correct one. Suppose you have the miter, the bevel, and the corner type all dialed in correctly. You still have to decide how the physical piece of crown sits on the saw before you pull the trigger, and that decision is where first-timers freeze.
For most cuts the crown goes face-down and upside-down on the saw table, with the part that touches the ceiling resting against the fence and the part that touches the wall flat on the table. This "nested" or "upside-down and backwards" method lets the saw's flat table and vertical fence stand in for the wall and ceiling, so a compound cut becomes manageable. Get that orientation backwards and the correct angles produce a piece that's beveled the wrong direction, useless, with the cut on the wrong face. The trade shorthand is blunt for a reason: crown face-down on the table, spring against the fence, and know whether you're keeping the left piece or the right. A diagram that shows which way the piece faces is worth more at the saw than another decimal place on the bevel.
If you want a tool that takes your spring angle and your measured corner, runs the compound-miter trig live for both inside and outside, and draws the saw orientation so you can see which face goes down before you cut, Crownset does exactly that, free and on a phone. It's most useful in the moment right before the cut, when the math is settled and the only question left is which way to hold the piece.