Take one part and print it at three layer heights. Here is the number that should decide how you think about the setting:
The amount of plastic does not change.
| Layer height | Layers | Plastic extruded | Time |
|---|---|---|---|
| 0.1 mm | 800 | 82,224 mm³ | 6.82 h |
| 0.2 mm | 400 | 82,224 mm³ | 3.41 h |
| 0.3 mm | 267 | 82,847 mm³ | 2.29 h |
The part is an upright 80 mm tall with 3,600 mm² of footprint, three walls and 20% infill, at 100 mm/s on a 0.4 mm nozzle. You can re-run it with your own geometry.
Identical volume at 0.1 and 0.2, and only a whisker more at 0.3 — that last one moves because the solid top and bottom skins are specified as a thickness, so a coarser layer rounds up to a slightly thicker skin. Everything else is the same plastic in the same places.
So layer height buys you exactly one thing at exactly one price: surface quality, paid for in hours.
Does halving it exactly double the time?
In the model above, yes, to the second decimal — and that is worth treating with suspicion rather than satisfaction, because it is a consequence of how the model is built. It divides the plastic by a flow rate and adds a fixed allowance per layer. Halve the layer height and the flow per second halves too, so the extrusion time doubles cleanly.
A real slicer will usually report a little less than double, because some of what it is timing does not scale with layer count at all: the initial heat-up, the purge line, long travel moves between islands, and the acceleration limits that dominate short segments. It can also report more than double on a part covered in small features, where every layer pays a fresh set of accelerations and retractions.
The useful reading is therefore "close to double, and never anywhere near free". Whether halving layer height doubles print time has the fuller answer, and the estimator states its own ±30% band against a real slicer for exactly this reason.
Where that lands in money
Take those three prints in PLA at $20 a kilogram — 102 g, about 34.2 m off the reel in all three cases — on a machine amortised at 30 cents an hour, with an 8% failure rate.
| Layer height | Filament | Machine time | Total |
|---|---|---|---|
| 0.1 mm | $2.04 | $2.05 | $4.57 |
| 0.2 mm | $2.04 | $1.02 | $3.39 |
| 0.3 mm | $2.04 | $0.69 | $3.01 |
At 0.1 mm the machine time costs more than the material does. That is the crossover worth remembering: on a fine-layer print you are no longer mostly buying plastic, you are mostly buying hours — and if you count your own attention at all, the total shifts further still.
The pattern generalises. Anything that adds layers adds cost with no return in material, and anything that adds material at least gives you material.
The strength question, and how good the evidence actually is
The widespread belief is that thinner layers make stronger parts. The evidence does not support it, and it points gently the other way.
For load applied across the layers — the direction parts actually fail in — thicker layers give you fewer weld interfaces, and each weld carries more heat into the layer below because more molten plastic arrives at once. Independent community bench testing has repeatedly found thicker layers equal or slightly better in that direction. The site's own nozzle reference says the same thing about the 0.8 mm case: fewer, thicker welds usually make the part stronger, not weaker.
Two honest caveats on that.
It is community bench testing, not standards-lab work. Absolute figures vary between testers, between materials and between machines, and anyone quoting a single percentage improvement is over-reading their sample. What has been consistent is the direction and the fact that the effect is small either way.
And it is not the lever that matters. Layer height changes cross-layer strength by a little; wall count changes it by a lot, because walls are continuous extrusions running along the load rather than a stack of interfaces across it. If a part is breaking, how many walls it has is the first question and layer height is somewhere below fifth. The layer height and strength guide goes into the mechanism; what layer height to use for strength is the short answer, and it is not "the smallest one".
If your layers are separating outright, that is not a layer-height problem at all — it is temperature, cooling or moisture, and no setting on this page will fix it.
What layer height genuinely changes
Four things, all of them real:
- Curved and sloped surfaces. Stepping on a shallow slope is a function of layer height and slope angle, and it is the reason display prints get thin layers.
- Vertical feature resolution. An engraved detail a third of a millimetre deep cannot exist on 0.3 mm layers.
- Overhang and bridging quality. Thinner layers bridge and overhang more cleanly, because each unsupported extrusion is smaller and lighter.
- Seam visibility. More layers means more seams, but each is smaller. Which reads better depends on the part.
Notice that all four are appearance or fine-detail concerns. None of them is structural. That is the whole argument of this post: layer height is a finish setting that people have been taught to treat as a strength setting, and treating it that way costs hours on functional parts that will never be looked at closely.
The ceiling that stops you going coarser
You cannot simply keep raising it. Above roughly four fifths of the nozzle diameter, layer bonding becomes unreliable — the extrusion cannot spread and press into the layer below properly — and the estimator warns when a setting crosses that line. On a 0.4 mm nozzle that puts the practical maximum around 0.3 mm.
Which is the real reason to think about layer height and nozzle diameter together rather than separately. If you want coarser layers, the route is a bigger nozzle, not a braver setting: an 0.8 mm nozzle supports 0.6 mm layers comfortably and raises the flow ceiling to match.
Speed will not substitute. Past a certain point the hotend cannot melt plastic fast enough and the machine quietly slows itself down, which is what the volumetric flow ceiling means in practice and why raw speed disappoints people who have not changed anything else.
A policy that holds up
Print functional parts at 0.3 mm on a 0.4 mm nozzle and stop apologising for it. Reserve fine layers for the surfaces you will actually look at, and for parts with genuine Z detail. When something needs to be stronger, add walls before you touch the layer height — and add walls before you touch the infill, which is a bigger saving again. Then take a part you print regularly, run it through the cost calculator at both layer heights, and multiply the difference by however many of them you have made this year. That is what the habit has been worth.