This study evaluated the effect of offset on the precision of three-dimensional (3D)-printed splints, proposing to optimize the splint design to compensate for systematic errors. 14 resin model sets were scanned and offset as a whole by given distances (0.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, and 0.40mm). Intermediate splints (ISs) and final splints (FSs) were generated from the non-offset and offset models and grouped correspondingly, named as splint type-offset value, IS-0.05, for instance. Dentitions occluded with the splint were scanned. Translational and rotational deviations of the lower dentition relative to the upper dentition were 3D measured. Deviations of ISs and FSs were more evident in the vertical and pitch dimensions, and were mostly acceptable in other dimensions. ISs with offset ≥ 0.05mm showed vertical deviations significantly below 1mm (P < 0.05) while ISs with 0.10- to 0.30-mm offsets had pitch rotations significantly lower than 1° (P < 0.05). The Pitch of IS-0.35 was significantly larger than ISs with 0.15- to 0.30-mm offsets (P < 0.05). Meanwhile, FSs fit better as the offset increased and FSs with offsets ≥ 0.15mm all had deviations significantly lower than 1mm (for translation) or 1° (for rotation) (P < 0.05). Offset affects the precision of 3D-printed splints. Moderate offset values of 0.10 to 0.30mm are recommendable for ISs. Offset values ≥ 0.15mm are recommended for FSs in cases with stable final occlusion. This study found the optimal offset ranges for 3D-printed ISs and FSs via a standardized protocol.