PURPOSE. The purpose of this study was to compare the accuracy of the implant master cast according to the type (pick-up, transfer) and the length (long, short) of the impression copings. MATERIALS AND METHODS. The metal master cast was fabricated with three internal connection type implant analogs (Osstem GS III analog), embedded parallel and with $10^{circ}$ of mesial angulation to the center analog. Four types of impression coping were prepared with different combinations of types (transfer, pick-up) and lengths (long, short) of the coping. The impressions were made using vinyl polysiloxane (one step, heavy + light body) with an individual tray, and 10 impressions were made for each group. Eventually, 40 experimental casts were produced. Then, the difference in the distance between the master cast and the experimental cast were measured, and the error rate was determined. The analysis of variance was performed using the SPSS (v 12.0) program (${alpha}$= .05), and the statistical significance was set at P < .05. RESULTS. The ANOVA showed that the pick-up type impression coping exhibited a significantly lower error rate than the transfer type. However, no significant difference was observed with respect to the length of the impression coping. Additionally, no significant difference was observed between the parallel and mesial angulated groups. CONCLUSION. Within the limitations of this study, the pick-up type impression coping exhibited a more accurate implant master cast than the transfer type in parallel group. The accuracy of the implant master cast did not differ for different lengths of impression coping of at least 11 mm. Additionally, the accuracy of the implant cast was not different for the parallel and $10^{circ}$ mesial angulated groups.