The goal of this study was to evaluate the effect of using die spacers on the marginal fit of NiCr (M1) and NiCrBe (M2) alloys and commercially pure titanium (cpTi) (M3) copings cast by the lost wax technique. Using a metal matrix, 45 resin added extra hard type IV stone models were obtained for the fabrication of wax patterns under the following conditions: no die spacer (A), with one die spacer layer (B) and with two die spacer layers (C), with five repetitions for each condition (alloy x die). Each die was waxed and the wax patterns were invested as per manufacturer's instructions. Three wax patterns were embedded in each casting ring, each corresponding to one of the conditions. Each coping, seated to the metal matrix by a seating pressure standardizing device (SPSD), was taken to an optical microscope for measurement of marginal discrepancy. The obtained data (μm) were analyzed statistically by ANOVA and Tukey' test (α=5%). There was statistically significant difference (p<0.05) among the materials (M1=110.67; M2=130.33 and M3=148.33). Regarding the use of the die spacer, there was a statistically significant difference (p<0.05) among the three conditions (A=162.00; B=131.06 and C=96.67). It was concluded that there is less marginal discrepancy with two die spacer layers.