To compare the fit of 3D-printed titanium (Ti) and cobalt-chromium (Co-Cr) abutments with implants to computer numerical control (CNC)-milled, ready-made abutment-implant assemblies. Their clinical applicability was also evaluated by measuring removal torque values (RTVs) and percentage torque loss of abutment screws. A total of 138 abutments were included in the study: 92 abutments were fabricated with Ti and Co-Cr alloys using computer-assisted design (CAD) through selective laser melting, and 46 ready-made abutments were prepared. The fit of interfaces between 90 abutments from the three groups (30 ready-made, 30 3D-printed Ti, and 30 3D-printed Co-Cr abutments) and implant assemblies was demonstrated by scanning electron microscopy (SEM) and confocal scanning laser microscopy (CSLM). After 30-Ncm torque tightening of Ti abutment screws twice within 10 minutes, the RTVs and percentage torque loss of screws of 48 abutments (16 ready-made, 16 3D-printed Ti, and 16 3D-printed Co-Cr) were evaluated after 10 minutes of thermocycling and cyclic loading. The fits of 3D-printed Co-Cr abutments were not statistically different from those of ready-made abutments (P = .383), while the fit of 3D-printed Ti abutments was inadequate (P < .001). The RTVs of 3D-printed abutments after cyclic loading were significantly decreased compared with those of CNC-milled abutments (P < .001). The fit of interfaces between 3D-printed Co-Cr abutments and implants was adequate. The RTVs of 3D-printed Co-Cr abutments were not significantly different from those of CNC-milled abutments after 10 minutes of 30-Ncm torque tightening and thermocycling.