Statement of problemMicroleakage and loss of the composite resin sealing the screw-access channel are frequent complications of screw-retained implant-supported prostheses. How the screw-access channel should be best restored to reduce such complications is unclear. PurposeThe purpose of this in vitro study was to evaluate the microleakage and bond strength of 3 types of composite resins (flowable, packable, and bulk-fill) with or without a bonding agent treatment to seal the screw-access channel of 2 types of restorative materials (zirconia and Co-Cr alloy) with or without thermocycling. Material and methodsIn total, 240 yttria-stabilized tetragonal zirconia polycrystal (Y-TZP) specimens (IPS e.max ZirCAD) and 240 Co-Cr alloy (Vera PDS) specimens were prepared with a Ø3×3-mm cylindrical cavity at the center to simulate the screw-access channel. Three types of composite resins (flowable, packable, and bulk-fill resin) (Filtek Z350 XT Flowable Restorative, Filtek Z350 XT Universal Restorative, and Filtek One Bulk Fill Restorative) were applied to restore the access channel of the zirconia and Co-Cr specimens with or without a bonding agent (Single Bond Universal Adhesive). Microleakage and push-out bond strength were determined and compared by dividing the specimens into experimental groups with or without thermocycling (1000 times with 30 seconds at 5 ±2 °C and 55 ±2 °C). The results were analyzed by using a 1-way ANOVA and 4-way ANOVA. Adjustment for multiple comparisons was made with the Tukey Honestly Significant Difference (HSD) test. ResultsThe specimens subjected to thermocycling showed a lower bond strength (P<.001) and higher microleakage (P<.001) than specimens stored in a constant-temperature water bath. Specimens treated with bonding agents showed a higher bond strength (P<.001) and lower microleakage (P<.001) than specimens not treated with a bonding agent. Higher bond strengths were observed in the order of bulk-fill resin, packable resin, and flowable resin (P<.001). Packable resin showed higher microleakage than flowable resin and bulk-fill resin (P<.05). No significant difference in microleakage was found between the flowable resin and bulk-fill resin (P>.05). ConclusionsHigher bond strengths were observed in the order of bulk-fill resin, packable resin, and flowable resin. Less microleakage was observed in the flowable resin and bulk-fill resin than in the packable composite resin. Bonding agent treatment was effective in increasing bond strength and decreasing microleakage. Zirconia and Co-Cr showed a bond strength similar to that of composite resins, but zirconia showed higher microleakage than Co-Cr. Restoring the screw-access channel with the bulk-fill resin should increase bond strength and reduce microleakage.