3D printing or CNC? A decision tree for prototype managers

Geometry, quantity, mechanical requirements, surface finish, and lead time — the five questions that decide between additive and subtractive.

Question 1 — Can it be machined at all?

Internal lattices, conformal cooling channels, organic generative-design shapes, anatomical implants — these are 3D-only territory. CNC requires tool access, so anything with hidden internal geometry is impossible to machine. If your part has features a 4 mm tool can not reach, the answer is 3D.

If the part is prismatic (mostly flat surfaces, holes, pockets, edges), CNC will be cheaper, faster, and tighter tolerance.

Question 2 — What quantity?

1-10 parts: 3D printing wins almost always (no setup, no fixturing). 10-100 parts: depends on geometry — simple parts go CNC, complex stay 3D. 100-1000 parts: CNC dominant unless geometry forces 3D. 1000+: CNC for now, until carbon fiber composite 3D printing matures (a few years out for production volumes).

Question 3 — Mechanical requirements

DMLS metal parts approach 90% of wrought material strength after heat treatment — close enough for most structural applications. SLS PA12 nylon hits 80-85% of injection-molded nylon strength. SLA resin parts are weakest, fine for visual prototypes but not load-bearing.

If you need full wrought-material strength (aerospace structural, high-cycle fatigue parts), CNC from billet stock. Otherwise, modern 3D printing is genuinely production-grade.

Quick decision

Visual model only: SLA. Functional prototype with moving parts: SLS or MJF. Metal end-use part with complex geometry: DMLS. Metal end-use part with simple geometry: CNC. Anything in production volumes above 1000 units: CNC, then injection mold once volumes justify tooling.