Statement of problemStudies that evaluated the accuracy of different methods for making an impression of 2 implants in close proximity or with adverse axial convergence are lacking. PurposeThe purpose of this in vitro study was to compare the accuracy of impressions made with metal and resin computer-aided design and computer-aided manufacturing (CAD-CAM) impression copings and an altered cast technique. Material and methodsTwo implant analogs with an interimplant angulation of 50 degrees were placed in an acrylic resin master model. Thirty impressions were made with 3 techniques (n=10). Two angled abutments were scanned to print 40 impression copings from a castable resin, half of which were cast with nickel-chromium alloy (M-CAD) and half of these were used in the R-CAD group. Impressions were made and poured. For the altered cast technique (Alt-cast), casts with only the mesial analogs were made and hollowed at the site of the distal analogs. The position of the distal analogs was transferred by resin indices (n=10) fabricated with light-polymerized resin on the master model. The linear displacements of mesial (ΔR1) and distal (ΔR2) analogs and the interanalog distance (ΔRt) were measured with a coordinate measuring machine. The data were analyzed by using the Kruskal-Wallis and Dunn post hoc tests (α=.05). ResultsNo significant difference was found between the M-CAD and R-CAD groups regarding ΔR1 (P=.722), ΔR2 (P=.576), or ΔRt (P=.939). However, the Alt-cast group differed from the M-CAD group in ΔR1 (P<.001), ΔR2 (P<.001), and ΔRt (P=.002) and also from the R-CAD group in ΔR1 (P=.001), ΔR2 (P=.002), and ΔRt (P=.003). ConclusionsThe impressions made with metal and resin CAD-CAM impression copings were more accurate than those made with the Alt-cast technique although all techniques had acceptable accuracy.