PURPOSE. This study evaluated the effectiveness of various methods for removing provisional cement from implant abutments, and what effect these methods have on the retention of prosthesis during the definitive cementation. MATERIALS AND METHODS. Forty implant fixture analogues and abutments were embedded in resin blocks. Forty cast crowns were fabricated and divided into 4 groups each containing 10 implants. Group A was cemented directly with the definitive cement (Cem-Implant). The remainder were cemented with provisional cement (Temp-Bond NE), and classified according to the method for cleaning the abutments. Group B used a plastic curette and wet gauze, Group C used a rubber cup and pumice, and Group D used an airborne particle abrasion technique. The abutments were observed using a stereomicroscope after removing the provisional cement. The tensile bond strength was measured after the definitive cementation. Statistical analysis was performed using one-way analysis of variance test (${alpha}$=.05). RESULTS. Group B clearly showed provisional cement remaining, whereas the other groups showed almost no cement. Groups A and B showed a relatively smooth surface. More roughness was observed in Group C, and apparent roughness was noted in Group D. The tensile bond strength tests revealed Group D to have significantly the highest tensile bond strength followed in order by Groups C, A and B. CONCLUSION. A plastic curette and wet gauze alone cannot effectively remove the residual provisional cement on the abutment. The definitive retention increased when the abutments were treated with rubber cup/pumice or airborne particle abraded to remove the provisional cement.