Purpose: The purpose of this study was to evaluate the clinical efficacy of implant prostheses retained by screws and cement (SCPs) by examining the reverse torque values (RTVs) of the abutment screws and the marginal openings of the implant prostheses. Materials and Methods: Two implants (3.8 3 13 mm; Camlog Biotechnologies) were embedded in an acrylic resin block 5 mm apart. Eighteen copies of this resin specimen were fabricated and randomly divided into two groups. Two-unit implant prostheses with two different designs-purely cement-retained implant prostheses (group 1) and SCPs (group 2)-were made out of type IV gold alloy and placed on the implants. After tightening to about 30 Ncm, the preloading RTVs of the abutment screws were measured. After retightening the abutment screws or cementing the prostheses, followed by cyclic loading, the postloading RTVs of the abutment screws were examined. Also, the marginal openings of the prostheses in the two groups were measured under a stereomicroscope. These measurements were compared statistically. Results: The postloading RTVs and their differences from the preloading RTVs of the abutment screws demonstrated no significant differences between groups (P > .05). Group 2 prostheses showed significantly smaller marginal openings than group 1 prostheses (P .05). Conclusion: The forces generated when torquing the abutment screw of the SCP did not cause more loosening of the abutment screws than the purely cement-retained implant prosthesis. The SCP showed better marginal adaptation of the cement-retained part than the purely cement-retained implant prosthesis, possibly as a result of the screw-retained abutment seating the restoration. Within the limitations of this in vitro test, the SCP showed no significant difference in RTV of the abutment screw and a smaller marginal gap compared to a purely cement-retained implant prosthesis. Palabras clave: dental cement, dental screw, implant-supported prosthesis, marginal opening