Objectives : The aim of this study was to determine the accuracy of digitized stone models, impression materials compared to the master model and the reliability of the computer aided analysis. Methods : A master model(500B-1, Nissin dental product, Japan) with the prepared lower full arch tooth was used. Ten vinyl polysiloxane impressions(Examix$^{(R)}$, GC Industrial Corp, Japan) of master model were taken and type IV stone(aesthetic-base gold$^{(R)}$, Dentona, Germany) were poured in stone models. The linear distance between the reference points were measured and analyzed on the Delcam Copycad$^{(R)}$(Delcam plc, UK). The t-student test for paired samples was used for statistical analysis. Results : The mean differences to master model for stone model and impression material were 0.11~0.19mm, and 0.19~0.29mm, respectively. There were statistical differences in dimensional accuracy for full arch impression between master model and stone model/impressions(p<.05). Conclusions : Two different scanning methods showed clinically acceptable accuracy of full arch digital impression produced by them. These results will have to be confirmed in further clinical studies.