BACKGROUND: Three-dimensional (3D) printed models of the human skull and parts of it are being increasingly used for surgical education and customized preoperative planning. OBJECTIVE: This study, using the calvaria as a model, provides a methodologic analysis with regard to future investigations aimed at evaluating patient-specific skull replicas. METHODS: Postmortem computed tomography was used for 3D reconstruction of a skull. The digital model obtained was converted to a physical replica by 3D printing. This copy was compared qualitatively and quantitatively with the original, using both a classical anthropometric and a 3D surface scanning approach. RESULTS: Qualitatively, the replica and the original displayed good qualitative concordance. The quantitative deviations, as measured by osteometric tools, lay partly in the submillimetric area, partly between 1 and 2 mm. The maximum difference was 3.7 mm. On the basis of the surface scans, a mean deviation of 0.2930 mm (0.2677 mm) and a median difference of 0.2125 mm (0.0000e1.5509 mm) were observed for the inner surface. For the whole object, corresponding figures amounted to 0.9101 mm (0.5390 mm) and 0.8851 mm (0.000e3.2647 mm). - CONCLUSIONS: Qualitatively flawless replicas of the skull region investigated are feasible, subject to extensive manual CT image editing. However, neurosurgeons should be aware that models of one and the same patient will vary according to the production chain used by the 3D printing laboratory in charge. Methodologically, both classic anthropological and light-stripe-based comparisons are justified for use in future studies. For trials aimed at assessing mean deviations and topographic distribution patterns, optical 3D scanning technologies can be recommended.