The purpose of this study was to evaluate the dimensional accuracy of different materials used in the confection of dies. Two stainless steel standard models were confected. One of the models, which was 2 mm larger than the other model, was used to provide a uniform relief for the two-step putty-wash impression technique. Thirty impressions were obtained using a polyvinyl siloxane impression material and randomly divided into three groups (n = 10) according to the type of casting material: type IV dental stone, commercially available epoxy resin (Tri-Epoxy), and industrial epoxy resin (Sikadur). After the setting/polymerization of the casting material, the dimensional stability was measured in terms of the height, diameter of the base and diameter of the top from the obtained dies and from the standard metal model using a profile projector. Results were analyzed by ANOVA and Dunnet test (α = 0.05). In the height values, no significant difference was observed between the groups, except for Sikadur casts, which showed lower mean values. The Tri-Epoxi group showed statistically lower mean base diameter values, compared with the other groups, and both epoxy resin groups showed statistically lower mean top diameter values, compared with that for the type IV dental stone group. We concluded that type IV gypsum and the commercially available epoxy resin showed similar behavior in most areas. The industrial epoxy resin did not show the same characteristics, although the diameter of the base obtained with it was similar to that obtained with type IV dental stone.