A printed enclosure can look excellent on the build plate and still fail where it matters: a lid will not snap shut, a bearing will not seat, or two mating halves will sit slightly proud. That is why the question, how accurate are printed parts, needs a more useful answer than a single tolerance number. Accuracy depends on the printing process, material, part geometry, orientation, post-processing, and how the feature will be measured and used.
For startups and small product businesses, the practical goal is not chasing the smallest possible number on a spec sheet. It is designing a part that fits, functions, and can be made consistently at the quantity you need.
How Accurate Are Printed Parts in Practice?
Professional 3D printing can produce highly accurate functional parts, but it is not the same as machining every surface from a solid block. A well-controlled printed part may commonly hold roughly ±0.1 mm to ±0.3 mm on many features, with capability varying by process, material, machine calibration, and part size. Larger dimensions generally allow more total variation because small effects accumulate across the part.
That range is a starting point, not a blanket promise. A flat 20 mm feature, a thin vertical wall, a 150 mm enclosure, and a precision press-fit hole do not behave the same way. Printing introduces heat, curing, shrinkage, layer-based geometry, and support-related effects. Good manufacturing practice accounts for those realities before the first part is built.
It also helps to separate three terms that are often used interchangeably:
Accuracy is how close the finished part is to its intended CAD dimension. Repeatability is how consistently multiple parts come out alike. Resolution refers to the smallest detail a machine can create, such as layer height or pixel size. A printer can have fine resolution without automatically delivering tight overall accuracy.
For a prototype that needs to prove shape and general function, moderate tolerance may be perfectly acceptable. For a latching consumer product, a fluid path, an alignment fixture, or a mating assembly, the critical dimensions deserve a more deliberate plan.
The Process Has a Major Effect on Tolerance
FDM: capable, durable, and geometry-sensitive
Fused deposition modeling, or FDM, builds parts by laying down melted thermoplastic filament. It is an effective choice for durable prototypes, fixtures, housings, and low-volume functional components. Materials such as PLA, PETG, ABS, ASA, nylon, and reinforced filaments offer very different mechanical and thermal properties.
FDM accuracy is influenced heavily by thermal movement. As plastic cools, it contracts. A long, thin part may curl at the corners, while a box-like part can pull inward or outward depending on wall layout, infill, material, and orientation. Layer lines also affect round holes, angled surfaces, and small details.
For many FDM applications, designing about 0.2 mm to 0.4 mm of clearance between mating features is a reasonable early guideline. The correct allowance can be smaller or larger based on the material and fit type. A loose cover, a sliding rail, a snap feature, and a threaded connection each need different treatment.
SLA: fine detail with post-cure considerations
Stereolithography, or SLA, uses light to cure liquid resin. It is often selected when surface finish, fine features, and detailed presentation models matter. It can also produce useful functional prototypes with engineering resins chosen for toughness, heat resistance, flexibility, or other requirements.
SLA can achieve excellent detail, especially on small features. Still, resin parts require washing and post-curing, and those steps can introduce slight dimensional change. Support removal may leave marks, and thin sections can distort if the part is not oriented and supported correctly.
SLA is a strong option for intricate cosmetic parts, small mechanisms, detailed patterns, and fit-check prototypes. When a part requires repeated flexing, outdoor exposure, or high impact resistance, material selection matters as much as dimensional accuracy.
SLS: strong production-ready parts with a textured finish
Selective laser sintering, or SLS, fuses nylon powder layer by layer. The surrounding powder supports the part during printing, so complex shapes can be produced without the support structures common in FDM and SLA. That makes SLS particularly useful for internal channels, organic forms, interlocking assemblies, and batches of functional parts.
SLS nylon parts are tough and well suited to low-volume production, but powder-based printing has its own dimensional behavior. Cooling, part placement in the build, feature thickness, and finishing can all affect the final result. Surface texture is also rougher than a polished molded or machined part, which matters for sliding interfaces and sealing surfaces.
A good SLS design uses appropriate clearances, avoids unnecessarily thick solid masses, and identifies critical dimensions early. When needed, printed geometry can be paired with hardware, inserts, reamed holes, or secondary operations to achieve a more controlled functional interface.
Part Design Often Matters More Than the Printer
A capable printer cannot fully compensate for a part designed without manufacturing allowances. The most reliable projects begin by identifying which dimensions are truly critical. Usually, only a small portion of a part needs tight control: a mounting-hole pattern, a gasket channel, a latch interface, a shaft bore, or the gap between two assembly halves.
External dimensions and internal openings respond differently to printing. Holes often print undersized, especially in FDM, because the deposited bead has width and circular geometry is approximated in layers. Designing a hole at its nominal fastener size may produce a fit that is too tight. Depending on the process and application, it may be better to oversize the modeled hole, use a pilot hole for drilling, or specify a post-print reaming operation.
Wall thickness deserves the same attention. Very thin walls may be technically printable but inconsistent, while thick sections can retain heat or cure unevenly. Ribs, fillets, and thoughtful transitions help reduce stress concentration and warping. For flat components, small design changes such as chamfers, edge relief, or a different print orientation can make a meaningful difference.
Assemblies should be designed as assemblies, not as individual parts that happen to touch in CAD. Build in clearance for the selected process, account for surface texture, and test the actual interaction. A hinge that rotates freely in a digital model may bind once material thickness, layer texture, and post-processing are involved.
Orientation and Finishing Change the Result
The same CAD file can produce different results when printed in a different orientation. On FDM, a vertical cylinder may create a more circular outside diameter than the same cylinder printed horizontally, while the horizontal version may be stronger in another direction. On SLA, orientation affects support placement, drainage, and the risk of distortion during curing. On SLS, build placement and thermal history influence shrinkage behavior.
Finishing is another variable. Sanding improves appearance but removes material. Vapor smoothing can change surface behavior and small dimensions. Dyeing, coating, bead blasting, and machining all have their place, but each should be considered before setting final tolerances.
This is why a dimensional requirement should include more than a number. It should identify the feature, the allowable variation, the measurement method, the mating condition, and any planned finishing steps. A requirement such as “must fit a 6 mm pin with light hand pressure” gives an engineering team far more useful direction than simply calling out a 6 mm hole.
A Practical Path to Better-Fitting Printed Parts
For a new product, the fastest path is often a staged approach. Start with a fit-check prototype to validate overall geometry, hand feel, assembly sequence, and obvious interference points. Then test the critical interfaces with the intended material and process. Finally, lock the design and document the features that must be checked during production.
This approach is especially valuable before ordering a batch. A small adjustment to a latch gap or hole size during prototyping is inexpensive. Finding that issue after packaging or customer shipment is not.
At Tech Connext, the work starts by reviewing how a part needs to function, not just whether it can be printed. That may mean recommending a different material, changing a wall transition, adding clearance at a mating feature, or planning a secondary operation where it delivers real value. The objective is a part you can hold, test, and trust in the next stage of development.
If your product depends on a specific fit, bring that requirement into the conversation early. A physical test part made under the intended process will answer more than a tolerance chart ever can.