To three-dimensionally evaluate the internal fit at the implant-abutment interface of abutments fabricated with different workflows using a combination of the silicone replica technique and microcomputed tomography (μCT).Thirty abutments were fabricated to restore internal-connection implants and were divided into three groups according to fabrication method: (1) full digital (abutment machined using CAD/CAM system); (2) Ti-Base (prefabricated standard Ti-Base abutments); and (3) UCLA (UCLA-type abutments) (n = 10/group). Linear and volume measurements were performed to assess the internal misfit using a silicone replica of the implant-abutment interface misfit area, which was three-dimensionally reconstructed after μCT. The internal discrepancies in three different regions of interest (Gapsuperior, Gapmarginal, and Gapcenter) were assessed. Data were statistically evaluated using ANOVA and Tukey test (P < .05).Ti-Base and UCLA abutments presented significantly lower misfit volume (0.49 mm3, 95% CI: ± 0.045 mm3 and 0.48 mm3, 95% CI: ± 0.045 mm3, respectively) and mean internal gap (25.20 μm, 95% CI: ± 3.14 μm and 27.97 μm, 95% CI: ± 3.14 μm, respectively) than the full digital group (0.70 mm3, 95% CI: ± 0.045 mm3; 34.90 μm, 95% CI: ± 3.14 μm) (P < .001), but did not differ from each other (P = .825). While Gapcenter was significantly higher in the full digital group (P < .001), Gapsuperior and Gapmarginal did not demonstrate significant differences among groups. All regions were statistically similar within groups, except for Gapcenter in the full digital group, which exhibited higher mean values compared to the other regions (P = .000). The 3D measurements for quantification of internal discrepancy were strongly associated with the 2D measurements.Ti-Base and UCLA abutments exhibited better internal fit at the implant-abutment interfaces compared to a fully digitalized workflow (CAD/CAM custom abutments).