Purpose: The objective of this study was to investigate the reproducibility of milled acrylic resin and zirconia models using three-dimensional analysis. Material and Methods: Zirconia discs and acrylic resin blocks were milled to make models. Each model was scanned with a model scanner. The zirconia models were superimposed with other zirconia model, and the acrylic resin models were superimposed with other acrylic resin model threedimensionally. Root-mean-square (RMS) values were obtained. Distances between two superimposed models were measured. Statistical significances were tested using the Mann-Whitney U test and t-test. Results: The RMS values of the zirconia model and the acrylic resin model were not significantly different ( p ≥0.05). At the premolar site, distances between the zirconia models at the upper axial and at the lower axial wall were significantly smaller than for the acrylic resin models ( p <0.05). At the molar site, distance between the zirconia models at the lower axial wall was significantly smaller than for the acrylic resin models ( p <0.05). Conclusion: Although there is a difference in reproducibility between the milled acrylic resin model and the zirconia model in the localized region, no significant difference was found in the whole model.