To consistently provide passively fitting implant superstructures, an understanding of the accuracy and precision of all phases of fabrication and connection is required. The initial phase of fabrication, ie, impression making and cast forming, was investigated in an earlier report for a mandibular five-implant model. The current study evaluates the accuracy of working casts produced from impressions using two different transfer copings in a 15-degree divergent two-implant posterior mandibular model. While the indirect method is less cumbersome to use, it was found to be less accurate in the prior study. The purpose of this study was to see if the direct method is more precise for this clinical situation. A transfer was deemed effective in producing experimental casts if distances between specified points on the cast agreed with the corresponding distances on the master cast. The absolute value of the difference in distances between experimental and master casts was compared for the two techniques (two-sample t tests). No significant differences were noted (P > .05), and the power of the tests ranged from 0.70 to 0.96 against the one-sided hypothesis that the direct method had a smaller mean absolute difference in distance than the indirect method. This suggests no clear advantage in using the direct method in similar clinical situations. Comparison of these findings to other impression accuracy studies is made. Keywords: direct coping, impression transfer method, indirect coping, partially edentulous