To determine the accuracy (trueness and precision) of two different rapid prototyping (RP) techniques for the physical reproduction of three-dimensional (3D) digital orthodontic study casts, a comparative assessment using two 3D STL files of two different maxillary dentitions (two cases) as a reference was accomplished. Five RP replicas per case were fabricated using both stereolithography (SLA) and the PolyJet system. The 20 reproduced casts were digitized with a highly accurate reference scanner, and surface superimpositions were performed. Precision was measured by superimposing the digitized replicas within each case with themselves. Superimposing the digitized replicas with the corresponding STL reference files assessed trueness. Statistical significance between the two tested RP procedures was evaluated with independent-sample t-tests (P < .05). The SLA and PolyJet replicas showed statistically significant differences for trueness and precision. The precision of both tested RP systems was high, with mean deviations in stereolithographic models of 23 (±6) μm and in PolyJet replicas of 46 (±13) μm. The mean deviation for trueness in stereolithographic replicas was 109 (±4) μm, while in PolyJet replicas, it was 66 (±14) μm. Comparing the STL reference files, the PolyJet replicas showed higher trueness than the SLA models. But the precision measurements favored the SLA technique. The dimensional errors observed in this study were a maximum of 127 μm. In the present study, both types of reproduced digital orthodontic models are suitable for diagnostics and treatment planning.