Microgaps at the implant-abutment interface allow for microbial colonization, which can lead to peri-implant tissue inflammation. This study sought to determine the marginal accuracy of three different implant-zirconium oxide (zirconia) abutment configurations and one implant?titanium abutment configuration.Three combinations of implants with custom-made zirconia abutments were analyzed (n = 5/group): NobelProcera abutments/titanium inserts on Replace Select Tapered TiUnite implants (Nobel Biocare) (NP); Encode abutments/NanoTite Tapered Certain implants (Biomet 3i) (B3i); Astra Tech Dental Atlantis abutments/Biomet 3i NanoTite Tapered Certain implants (At). Five custom-made Encode titanium abutments/NanoTite Tapered Certain implants (Ti) were used as a control group. All abutments were fabricated with computer-aided design/computer-assisted manufacture. One-hundred twenty vertical gap measurements were made per sample using scanning electron microscopy (15 scans x 4 aspects of each specimen [buccal, mesial, palatal, distal] x 2 measurements). Analysis of variance was used to compare the marginal fit values among the four groups, the specimens within each group, and the four aspects of each specimen.Mean (± standard deviation) gap values were 8.4 ± 5.6 Μm (NP), 5.7 ± 1.9 Μm (B3i), 11.8 ± 2.6 Μm (At), and 1.6 ± 0.5 Μm (Ti). A significant difference was found between B3i and At. No difference resulted between NP with the other two groups. Gap values were significantly smaller for Ti relative to all zirconia systems. For each ceramic abutment configuration, the fit was significantly different among the five specimens. For 12 of the 15 ceramic abutment specimens, gap values sorted by aspect were significantly different.The implant?titanium abutment connection showed significantly better fit than all implant?zirconia abutment configurations, which demonstrated mean gaps that were approximately three to seven times larger than those in the titanium abutment system.