Yes, 3D printed parts are fully legal in FRC. No rule singles out additive manufacturing as restricted: a printed bracket is treated exactly like a CNC'd one, under the same FABRICATED ITEM rules covering anything "altered, built, cast, constructed, cut, machined, manufactured, [or] modified... into the final form in which it will be used on the ROBOT." What trips teams up isn't whether printing is allowed, it's when you can run the printer and when the design was made, both covered below, along with the part that actually decides matches: picking a material and settings so the thing you printed doesn't shatter in Q47.
What the rules actually say
A 3D printed part is a FABRICATED ITEM, not a COTS item, the moment you print it. That puts it under R301, the $600 Fair Market Value cap: for fabricated parts, FMV is "the value of the material and/or labor, except for labor provided by team members." So the FMV of a part you print yourself is just the filament cost, since your own labor is free. A bracket printed with $2 of PETG has a $2 FMV no matter how many hours it took to design, and that almost never touches the cap. Rule numbers and that $600 figure are current as of this manual: FIRST renumbers rules and adjusts caps most seasons, so confirm both against the current manual before you lean on them.
Two rules matter more in practice:
Pre-Kickoff designs. Per R303, "ROBOT software and designs created before Kickoff are only permitted if the source files... are available publicly prior to Kickoff." The manual's own example is exactly the 3D printing trap: a team reuses a transmission "built from detailed designs developed prior to Kickoff," even though the physical part was built during the season, and that's a violation. Got a favorite Limelight mount designed last summer and never posted publicly? Redesign it after Kickoff, or publish it beforehand to clear yourself to reuse it as-is.
Overnight printing at events. Per E510, "Teams may not run a 3D printer or other automated manufacturing process overnight in the pit." That's a flat ban on the hours the pit is closed, staffed or not, not just a rule against leaving one unattended: print all day while the pit is open, then shut it off. E401 does exempt "3D printed parts" from the Load-In lockout that otherwise keeps new items out after the deadline, so parts printed off-site overnight can still come in the next morning.
One more thing worth knowing: bumper backing has no mandated material. R402 just requires a backer "at least 4.5in tall which supports the padding," no wood required in the current text (older seasons did specify plywood, per the Bumper Guide). A thick-walled printed backer is technically legal, though plywood stays cheaper and faster to cut. See our bumpers guide for the rest, and check the current manual before assuming a material spec: these rules get tweaked most seasons.
What's actually worth printing versus machining
3D printing wins when a part is complex to machine, doesn't see much load, or needs to exist five minutes from now. It loses badly on anything absorbing repeated hits or transmitting real torque.
Good candidates for printing:
Sensor and camera mounts (Limelight, PhotonVision cages, beam-break brackets)
Wire looms, strain relief, and low-load electronics brackets or standoffs
Idler pulley or bearing covers, dust shields, chain guards
Rapid-iteration prototypes and one-off jigs or fixtures
Better machined, cut, or bought COTS:
Primary drivetrain and chassis structure, where aluminum holds stiffness under repeated collisions that FDM plastic can't match at comparable weight
Gearbox side plates carrying bearing loads
Motor pinions and high-torque drive gears. FRCDesign.org's design handbook puts it plainly: "3D printed gears must be used with caution in high torque applications like motor pinions or drive gears," since all the load runs through a handful of meshing teeth at once.
Swerve module drivetrain gearing. Even MAXSwerve, REV's budget module, ships with "all steel gears in [the] drive powertrain," per REV's own spec sheet; see swerve modules compared.
PLA: the default on most printers, and worth avoiding for anything structural. Stiff but brittle, and it softens at temperatures a robot can see in a hot pit or a car trunk. Fine for jigs and anything that will never take an impact.
PETG: the reasonable default for most functional FRC parts. One team on a Chief Delphi pulleys thread put it plainly: "our team prefers PETG pulleys and will only use PLA ones if there is no other option." FRCDesign.org separately notes PETG offers "significant improvement in impact strength and tensile strength over PLA," and a peer-reviewed compression test of FDM columns found a PETG variant held about 4,077 N before failure versus about 2,720 N for PLA, which failed brittle at the lower load.
ABS: tougher and more heat-resistant than PLA, but warps without a heated bed and enclosure, and layer adhesion suffers near open windows or fans. Workable if your printer's already dialed in for it, not worth fighting otherwise.
Nylon, including carbon-fiber-reinforced variants like PA-CF: the toughest common option. One team on Chief Delphi recommends Nylon 910 specifically for printed gears, calling it "fairly wear resistant." Abrasion-resistant, but plain nylon is hygroscopic: it pulls moisture from the air and prints weak, stringy layers if it isn't dried first.
Polycarbonate: per FRCDesign.org, "nearly double the tensile strength and bending strength of PLA... when properly annealed." Strong, but needs a high hotend temperature and usually an enclosure, a step up most teams don't need outside a genuinely high-stress bracket.
TPU: a flexible, rubber-like filament FRCDesign.org calls out for sensor mounts taking repeated impacts, since it bends and absorbs energy instead of cracking.
Wall and infill settings that survive a match
Print orientation matters more than any slicer setting. FDM parts are strongest in the X-Y plane and weakest between layers: per FRCDesign.org, "prints are typically strongest in the x-y plane of the printer as in the z direction the layers can delaminate." Orient the part so load runs in-plane, across a layer, not through the layer stack where it can peel the layers apart.
For anything carrying real force:
Increase wall (perimeter) count and infill well above the slicer defaults. FRCDesign.org notes power-transmission parts "may need to be printed with more walls and higher infill percentage," and some designs need close to 100% infill. Real practice backs this up: on a Chief Delphi pulley thread, one team runs 100% infill for pulleys in "high rpm (shooter)" and "high torque (pivot)" subsystems specifically because of the load, and a peer-reviewed compression test of PETG honeycomb structures similarly found 100% infill printed upright gave the best energy absorption before failure.
Reinforce any hex bore a shaft rides in. Repeated torque wears a printed hex hole round, exactly what another team on that thread describes: "due to the excessive amount of torque, our hex bores started to be ruined," fixed with metal hex inserts. A heat-set brass insert or oversizing to a larger hex adapter works the same way.
Design holes as teardrops, not circles, when printed flat (parallel to the bed): a true circle needs the top arc to bridge across empty space before it's supported, so it sags undersized. FRCDesign.org recommends a teardrop with roughly 100 degrees at the top instead. For press-fit tolerances, plan on 0.004 to 0.020 in (0.1 to 0.5 mm) and test on your own printer, since fit varies machine to machine.
The classic failures
Delamination under shock loads. A part solid under steady load can pop apart along a layer line the first time it takes a hard hit, since layer bonds are weaker than the plastic. Often an orientation problem more than a material one: reprint with the load running in-plane, not through the layer stack.
Elephant's foot ruining a press fit. The first layer of a print often squishes wider than the layers above it as it presses onto the bed, exactly the wrong place for that on a bore that needs a tight fit. One team's fix, from that same Chief Delphi thread: a small chamfer, about 0.03 in, on the bore's edge, which prevents the squish and eases the press fit.
Gear teeth stripping under load. Printed teeth concentrate all the transmitted force into whichever two or three are in mesh at once, unlike a timing belt, which spreads force across roughly half the teeth on a pulley. One team's report on Chief Delphi is typical: nylon gears on their intake had "teeth [that] have started shearing off in testing," exactly why motor pinions and drive gears belong on the "better machined" list above.
Warping and layer shift. ABS and other high-shrink materials warp off the bed without a heated enclosure. Separately, a loose belt or bumped gantry mid-print shows up as a visible layer shift, a calibration issue rather than a material one, but it produces a part that looks fine and fails immediately under load.
None of this is exotic. Print what makes sense, keep real torque on the machine shop list, pick PETG or nylon over PLA once something has to survive a hit, and orient for strength over print time. That's most of what separates a bracket that lasts a season from one that's back on the printer before quals are over.
Sources & corrections
This article is AI-assisted: drafted from primary sources, then reviewed and edited by hand. Errors still get through. When one is reported we fix it and write down what changed — publicly, in the corrections log.
You came here for one answer. These lessons teach the same subject properly, in the order a team actually learns it. All 35 are free and none of them need an account to read.