Quick Answer: Kerf is the width of material a laser beam removes as it cuts, which makes every finished piece slightly smaller than the shape you designed. Diode lasers typically cut a kerf of roughly 0.1-0.4mm and CO2 lasers roughly 0.07-0.5mm, with hobby CO2 machines commonly landing around 0.16mm in 3mm plywood, according to community-tested kerf reference data and laser-forum measurements. You measure your own kerf by cutting a test square or comb pattern and comparing the result to the designed size, then enter half that value as the “Kerf Offset” in LightBurn so interlocking parts like finger joints come out at the size you actually drew.
If you’ve ever designed a box with finger joints that measured perfectly in software and then wouldn’t fit together once cut, kerf is almost certainly why. It’s an easy thing to ignore for years of pure engraving work, and then it becomes the first thing you have to understand the moment you try to cut anything that needs to interlock, slot together, or match another piece exactly.
What is kerf, exactly?
Kerf is the width of material the laser beam actually vaporizes as it travels along a cut path — not a design flaw, just physics. A laser beam isn’t infinitely thin; it has a real width at the point it focuses on your material, and that width of material disappears as smoke and vapor with every pass. A shape drawn at exactly 100mm wide in your design software comes out slightly smaller once it’s actually cut, because the beam removed a sliver of material all the way around the outline.
This matters far more for cutting than for engraving, since engraving only marks a surface rather than removing a full-thickness strip of material along a path.
Typical kerf widths by laser type
| Laser type | Typical kerf | Notes |
|---|---|---|
| Diode laser | ~0.1-0.4mm | Wider and less consistent due to a rectangular, not circular, beam spot |
| CO2 laser (hobby, 3mm material) | ~0.16mm | Fairly consistent across most hobby-class CO2 machines at this thickness |
| CO2 laser (general range) | ~0.07-0.5mm | Varies with material, thickness, focus, and power/speed settings |
| CO2 laser cutting 2mm mild steel | ~0.30-0.45mm | Metal cutting runs a noticeably wider kerf than wood or acrylic |
Kerf isn’t a fixed spec you can look up once and reuse forever — it shifts with material, thickness, focus height, power, speed, and whether air assist is running, which is why every serious kerf guide tells you to test it on your own machine rather than trust a published number. If you haven’t dialed in your power and speed settings yet, start with our laser engraver settings cheat sheet first, since a cut that’s under-powered or too fast changes the kerf along with the cut quality.
Diode lasers: watch for directional differences too
Most diode modules produce a rectangular rather than perfectly circular beam spot, which means the kerf on a horizontal cut can measure differently than the kerf on a vertical cut on the exact same machine — typically by around 0.05-0.10mm on common home diode lasers. That’s one reason diode kerf runs wider and less predictable overall (roughly 0.15-0.50mm) than CO2 kerf, and why a single offset number is always a best-fit approximation rather than a perfect match on every edge of a part. If you’re deciding between laser types in the first place, our diode vs CO2 laser comparison covers this precision gap alongside cost and material limits.
How to measure your own kerf
The square method (fastest)
Cut a 100mm square from your material and measure the actual result with digital calipers. If it comes out at 99.7mm, your kerf is 0.3mm. This is the quickest test and good enough for most one-off projects.
The comb method (most precise)
Cut a frame with 20 thin strips inside it — a comb pattern — creating 20 individual cuts across the piece. Push the strips together at one end and measure the resulting gap at the other end, then divide that gap by 20. Averaging across 20 cuts smooths out any single measurement error and gives a more reliable number for projects where fit really matters, like finger-joint boxes or gears.
The stack method
Cut a stack of 10 identical strips from the same material and sheet. Measure the total width of the stack, subtract that from what 10 strips should measure at your drawn width, and divide the difference by 10 to get your average per-cut kerf.
Whichever method you use, test on the exact material, thickness, and settings you’ll actually use for the finished project — a kerf measured in 3mm plywood won’t carry over accurately to 6mm acrylic on the same machine.
Need a decent set of digital calipers to run these tests? Check current prices on Amazon →. Running a growing engraving shop and ordering test material or tools in bulk? A free Amazon Business account unlocks quantity discounts and tax-exempt purchasing on repeat supply orders.
Setting kerf offset in LightBurn
Once you know your kerf, open the Cut Settings Editor for the relevant cut layer in LightBurn and enter half your measured kerf value in the “Kerf Offset” field — per LightBurn’s own settings documentation, a measured 0.15mm kerf gets entered as 0.075mm. LightBurn then automatically shifts the toolpath inward on outside-contour cuts and outward on inside-contour cuts, so a part designed to be exactly 50mm actually cuts at 50mm instead of 49.7mm. This matters most for anything with interlocking pieces — finger-joint boxes, gears, and puzzle-style inlays — where a fraction of a millimeter of gap is the difference between a snug fit and a part that rattles or won’t seat at all. If you’re still choosing between laser control software, our LightBurn vs LaserGRBL comparison and what is LightBurn primer cover kerf offset support alongside the rest of each program’s feature set.
The bottom line
Kerf is simply the width of material your laser removes as it cuts — typically 0.1-0.4mm on a diode laser and 0.07-0.5mm on a CO2 laser, with hobby CO2 machines commonly landing around 0.16mm in 3mm plywood — and it makes every cut piece slightly smaller than its design. Measure yours with a quick square test or a more precise 20-strip comb test, then enter half that value as the Kerf Offset in LightBurn so interlocking parts like finger joints and gears fit the way you actually designed them. It’s a five-minute test that only matters once you start cutting parts that need to fit together — but the first time a box joint won’t close, it’s almost always the reason why.