The standard dimensional tolerance in FDM printing is ±0.2–0.5 mm in the X/Y axes and ±0.1–0.3 mm in the Z axis — which is why a model that measures a perfect 10.00 mm in CAD can come out at 9.7 mm or 10.3 mm after printing. The difference sounds small, but in a press fit or a bearing hole it can decide whether the parts assemble at all. A stable filament diameter, like Porima PLA with a ±0.03 mm tolerance, is one of the conditions that make these numbers meaningful in the first place.
This guide shows where dimensional deviations come from, how to design loose, sliding and press fits, and how to compensate typical errors directly in your slicer — without guesswork and without a string of test prints.
Starting Values for Fits — Table
| Parameter | Value | Comment |
|---|---|---|
| Loose fit (free movement) | +0.3 to +0.5 mm | Parts don't touch, large clearance |
| Sliding fit (free rotation) | +0.2 to +0.3 mm | Shaft rotates without resistance |
| Standard fit (light resistance) | +0.1 to +0.2 mm | Assembles by hand, holds on its own |
| Light press fit | -0.1 to -0.2 mm | Pushed together with fingers |
| Strong press fit | -0.2 to -0.4 mm | Needs a rubber mallet or a vise |
| Hole compensation (horizontal expansion) | -0.10 to -0.20 mm | Printers default to undersized holes |
| Porima filament diameter tolerance | ±0.03 mm | Affects material flow stability |
| X/Y accuracy on a calibrated printer | ±0.1 to ±0.2 mm | Without an enclosure and calibration it gets worse |
| Z-axis accuracy | ±0.1 to ±0.3 mm | Directly tied to layer height |
| PLA shrinkage | 0.2–0.3% | The most predictable material |
| PETG shrinkage | 0.5–0.8% | Higher than PLA, needs to be accounted for |
| ABS shrinkage | up to 2% | Highest risk of warping |
You'll find the full print temperature table for each material on the temperature settings page — dimensional accuracy and stable temperature go hand in hand.
Why Dimensions “Drift” in 3D Printing
In FDM printing, a digital model doesn't land on the print bed at a perfect 1:1 scale. Before the nozzle extrudes the first layer, the software slices the solid into cross-sections, and each one is reproduced with an accuracy limited by the nozzle diameter, line width, and how the machine handles mechanical stress. On top of that comes shrinkage during cooling — hot plastic takes up more space than cold plastic, so the finished model always ends up slightly smaller than the STL file.
The biggest factor is the calibration of your specific printer: material flow, extruder e-steps, and X/Y axis geometry. If these parameters aren't set correctly, no table in this article will help — start with flow calibration, step by step, because a 5% flow error translates directly into wall thickness and hole diameter.
The second factor is the material itself. PLA shrinks the least and most predictably, which makes it a good starting point for learning fits. PETG and ABS need their own compensation values, since their shrinkage and cooling behavior differ noticeably.
Where Dimensional Errors in FDM Printing Come From
Shrinkage During Cooling
When molten filament leaves the nozzle, it's at 200–260 °C, depending on the material. As it cools to room temperature, the polymer chains contract, and the model loses between 0.2 and 2% of its volume — the exact value depends on the material (see the table above). Shrinkage is greatest across large flat surfaces and at corners in the XY plane, which is why long, straight edges “drift” more than small details.
Printers Undersizing Holes
FDM printers consistently print holes smaller than designed, while outer edges come out slightly larger. The reason is mechanical: the nozzle has a finite width, and on a round hole the material “encroaches” toward the center more than it does on a straight wall. In practice, a hole designed at 10.0 mm can come out at 9.7 mm. That's why almost every slicer has a dedicated hole compensation parameter, covered in the next section.
How to Design a Fit, Step by Step
Loose and Sliding Fits
If two parts need to move freely relative to each other — a sliding drawer, for example, or a part you want to remove easily — add 0.3 to 0.5 mm of clearance per side. For a sliding fit, where a part should rotate without resistance but without excessive play, 0.2 to 0.3 mm is enough. These values work for typical dimensions up to about 50 mm; for larger parts, add another 0.05–0.1 mm for every additional 50 mm, since error accumulates with size.
Standard Fit
This is the most commonly used fit in practice — parts assemble by hand but hold together on their own, without glue or screws. It works with a difference of 0.1 to 0.2 mm between the hole and the pin. If the parts turn out too loose after printing, reduce that value by 0.05 mm and print a test sample instead of the whole model — it saves material and time.
Press Fit
A press fit requires negative clearance, meaning the hole must be smaller than the pin. A light press, one you can assemble with your fingers, is -0.1 to -0.2 mm. A strong press, requiring a rubber mallet or a vise, is -0.2 to -0.4 mm — and it's worth remembering that too much interference can crack a thin-walled part, especially one printed in ABS or PETG. For parts that need to hold a metal pin or a bearing, the lower end of this range works best.
Dimensional Compensation in Your Slicer
Instead of recalculating dimensions in CAD every time, most slicers (PrusaSlicer, Cura, OrcaSlicer) have a dedicated “Horizontal expansion” or “XY Size Compensation” field. Entering -0.1 to -0.2 mm there enlarges all holes and shrinks all outer edges by a fixed value, without editing the model itself.
It's worth remembering that this setting applies globally to the whole model, so if a single file has both press-fit and loose-fit holes, it's better to compensate specific holes directly in CAD and leave the global slicer compensation close to zero. Another parameter worth checking is “Elephant foot compensation” — it corrects the squashed first layer, which also affects the accuracy of the model's bottom edge.
Calibration Test — How to Check Your Own Printer's Accuracy
Before you start designing fits seriously, print a simple calibration model: a 20×20×20 mm cube and a plate with a series of holes from 4 to 12 mm in 1 mm increments. Measure the finished print with digital calipers and record the difference between the designed and actual dimensions separately for X, Y, and each hole.
This one-time test gives you your own compensation values for your specific printer, nozzle, and material — far more accurate than averaged figures from a table. It's worth repeating after any major change: a new nozzle, a different material, or after recalibrating e-steps.
Which Filament to Choose When Accuracy Matters
PLA is the best choice for your first fit tests — it has the smallest and most predictable shrinkage among popular materials, making it easier to arrive at stable compensation values. Porima PLA prints within a narrow temperature window, which further limits dimensional variation between prints.
If a part needs to work at higher temperatures or under load, PETG or a technical material is the more sensible choice — at the cost of slightly higher shrinkage that needs to be accounted for in compensation. For precise mechanical parts, like bushings or guides, technical filament with higher stiffness and lower creep under constant load also works well.
Settings for Popular Configurations
| Configuration | Hole compensation | Note |
|---|---|---|
| Bowden | -0.15 to -0.25 mm | More play in the feed system, worth leaving a bit more margin |
| Direct drive | -0.10 to -0.15 mm | Shorter filament path, usually more stable flow |
| CoreXY in an enclosure | -0.10 to -0.20 mm | Stable ambient temperature limits shrinkage |
| Multi-color AMS / CFS system | -0.15 to -0.25 mm | Extra color-change pauses can increase hole undersizing |
Most Common Mistakes
Designing a fit with no clearance. A hole with exactly the same dimension as the pin almost never assembles in practice — always plan for at least 0.1 mm of difference.
Copying someone else's compensation values 1:1. The values in this article are a starting point, not a ready-made recipe — every printer and nozzle has its own systematic error.
Ignoring print orientation. A hole printed horizontally comes out differently than the same hole printed vertically, because the “elephant foot” effect on the bottom edge also comes into play.
Too much interference in a thin wall. A -0.3 mm press fit in a wall that's only one line thick ends in a crack, not a lasting joint.
Skipping the calibration test before a large model. Printing an entire enclosure without first checking the fit on a small sample is the most common cause of wasted material.
Changing material without recalibrating. Switching from PLA to PETG without updating hole compensation almost always results in too tight a fit, because PETG shrinks more.
Measuring while still warm. A model measured right off the bed is still shrinking — wait until it cools to room temperature before recording your measurement.
Frequently Asked Questions
What is the standard tolerance for FDM 3D printing?
On a well-calibrated printer, it falls within ±0.1–0.2 mm in the X/Y axes and ±0.1–0.3 mm in the Z axis. On a printer without calibration or an enclosure, the deviation can reach as much as ±0.5 mm.
How much clearance should I leave for a press fit?
A light press, one you can assemble with your fingers, is usually -0.1 to -0.2 mm of difference between the hole and the pin. A stronger press, requiring a tool, is -0.2 to -0.4 mm.
Why do holes in 3D prints come out too small?
It's the result of nozzle geometry and how material lays down on curves — round contours consistently “eat into” the diameter. Typical undersizing is 0.2–0.3 mm, which is why almost every slicer has a dedicated hole compensation parameter.
Does tolerance depend on the material (PLA, PETG, ABS)?
Yes, and quite noticeably. PLA shrinks the least (0.2–0.3%), PETG moderately (0.5–0.8%), and ABS the most, up to 2%, which means the same compensation won't work identically for all three materials.
How do I check my own printer's accuracy?
Print a calibration cube and a plate with holes of different diameters, measure them with calipers, and compare them to the designed dimensions. The difference you measure is your real compensation value going forward.
Summary
Tolerances in 3D printing aren't magic — they're a combination of material shrinkage, nozzle geometry, and the calibration of your specific printer, and all of that can be measured and repeated. Start with the values in this article's table, print a small calibration test, and adjust your slicer's hole compensation to your own results.
If you're just getting started and looking for a material with a stable diameter and predictable shrinkage, check out our PLA filament range — the best starting point for learning precise fits.