Scaling a 3D print is straightforward in any slicer: select your model, type in a percentage or target dimension, and the software resizes it. The real challenge is understanding what happens to wall thickness, tolerances, and structural integrity when you change the size, and knowing when to scale in a slicer versus going back to your CAD file.
Uniform vs. Non-Uniform Scaling
Every major slicer (Cura, PrusaSlicer, OrcaSlicer, Bambu Studio) lets you scale a model in two ways. Uniform scaling changes all three axes by the same percentage, so a 150% scale makes the model 1.5 times larger in X, Y, and Z simultaneously. The proportions stay identical. Non-uniform scaling lets you stretch or shrink each axis independently, which is useful when you need a part that’s longer but the same width.
The catch with non-uniform scaling is that it affects everything along that axis, not just the dimension you care about. Stretching a box along its length also stretches every internal feature in that direction: screw holes become ovals, wall thickness changes on the resized faces, and snap-fit latches distort. If your model has functional geometry like hinges, threads, or interlocking joints, non-uniform scaling will almost certainly break those features. In that case, you’re better off modifying the model in CAD where you can selectively change the dimensions that matter.
Calculating the Right Scale Factor
The formula is simple: divide your target dimension by the current dimension, then multiply by 100 to get a percentage. If your model is 40 mm tall and you need it to be 60 mm, that’s (60 ÷ 40) × 100 = 150%. This works in reverse too. To shrink a 100 mm part down to 75 mm, you’d scale to 75%.
If you’re converting between model scales (say, 1:24 to 1:12), divide the original scale ratio by the target scale ratio. Going from 1:24 to 1:12 means 24 ÷ 12 = 2, so you’d scale to 200%. Online scale converters can handle these conversions automatically if the math gets cumbersome, but the underlying principle is always the same ratio.
What Happens to Wall Thickness
When you scale a model down, every wall gets thinner by the same proportion. A part with 2 mm walls scaled to 50% now has 1 mm walls. That might still print fine, but it changes the structural behavior significantly. Thin walls in FDM printing are more prone to delamination under load or heat because there’s less surface area for layers to bond. In resin printing (SLA/DLP), sudden changes in wall thickness can cause warping or cracking, especially in larger parts.
Scaling up has the opposite problem. Walls get thicker, which increases print time and material use. A model designed to be lightweight at its original size can become unnecessarily heavy at 200%. If you’re scaling up significantly, consider hollowing the model or reducing infill to compensate. When scaling down, check that your walls remain above the minimum printable thickness for your nozzle size. For a standard 0.4 mm nozzle, you generally need walls of at least 0.8 mm (two perimeters) for a solid result.
How Scaling Affects Tolerances and Fit
This is where scaling gets tricky. If you have two parts designed to fit together with a 0.2 mm clearance gap, scaling both parts to 150% turns that gap into 0.3 mm. Scale them down to 50% and the gap shrinks to 0.1 mm, which might be too tight for your printer to resolve cleanly. FDM printers have more dimensional variability than resin printers, so they need larger clearances to begin with.
As a general guideline for interlocking parts, 0.2 mm of clearance works for small, thin components, while 0.4 mm is better for larger, chunkier assemblies. If you scale a model and the fit becomes too loose or too tight, you have two options: adjust the clearance in CAD and re-export, or use your slicer’s compensation settings to tweak the result.
Most modern slicers have an XY hole compensation setting that adjusts only the internal holes without changing the outer dimensions of the part. A positive value enlarges holes; a negative value shrinks them. This is useful when scaling down causes holes to print too small, which is common because the printer’s extrusion path rounds inward on curved features.
Compensating for Material Shrinkage
Some materials contract as they cool, which means your printed part ends up slightly smaller than the digital model. PLA shrinks very little, but ABS and nylon contract noticeably. Slicers like OrcaSlicer have a shrinkage compensation setting in the filament profile that scales the entire model by a small percentage (something like 100.5% for ABS) to offset this contraction.
The recommended workflow is to calibrate shrinkage compensation for your filament first, then fine-tune hole and contour compensation on top of that. If you’re scaling a model and switching materials at the same time, the combined effect of intentional scaling plus shrinkage can catch you off guard. Print a test piece at the new scale and material before committing to a long print.
When to Scale in CAD Instead of the Slicer
Slicer scaling works perfectly for simple geometry: figures, decorative items, solid brackets, or anything without functional mating features. You select the model, punch in your percentage, and slice. It’s fast and requires no CAD knowledge.
Go back to CAD when your model has threads, snap fits, hinges, or any feature that needs to stay at its original dimension while the rest of the part changes size. CAD lets you selectively resize the body of a part while keeping a screw hole at exactly M3, for instance. Slicer scaling can’t do this because it treats the entire STL as one block of geometry. If you only have an STL and no source CAD file, you can import it into a free tool like Fusion 360 or Blender, but editing mesh files is significantly more cumbersome than modifying parametric CAD models.
Verifying Accuracy After Scaling
A common method for checking dimensional accuracy is printing a 20 mm calibration cube and measuring it with calipers. If it comes out at 20.1 mm on one axis, you know your printer is slightly over-extruding or has a steps-per-millimeter value that needs adjustment in firmware. People adjust this value, reprint, and confirm the cube hits 20 mm.
Calibration cubes have a real limitation, though. They can’t reveal whether your printer’s axes are skewed, meaning the X and Y axes aren’t perfectly perpendicular to each other. A skewed printer can produce a cube that measures 20 mm on every side but is actually a parallelogram. For scaled prints where precise geometry matters, printing a larger test piece or a diagonal test pattern gives you more reliable information than a small cube alone.
Scaling Models Larger Than Your Build Plate
When you need a print bigger than your printer can handle, the solution is splitting the model into sections, printing them separately, and assembling them afterward. You can do this in CAD or with dedicated splitting tools built into some slicers.
There are two approaches to splitting. The first is adding alignment features like pins, slots, dovetails, or grooves directly into the cut surfaces. These make assembly much easier because the pieces self-align, but they require a printer accurate enough that the alignment features actually fit together. The second approach is simple straight cuts with no alignment features. Straight cuts are more forgiving if your printer has some dimensional variation or warping, but assembly takes more patience since you have to manually position each piece and hold it until the adhesive cures.
For either method, labeling each piece with a number or letter (built into the model or marked with a pen) saves frustration during assembly. For bonding, adhesive works well for decorative or display pieces. Functional parts that need to handle stress benefit from mechanical fasteners like embedded screw threads or bolt pockets designed into the split. If you’re printing the outer shell only and hollowing the interior, you can save substantial material and print time on large-format projects without sacrificing surface quality.