Objective: The cementation of fixed dental prostheses (FDPs) on implants involves the risk of undetected excess cement. If a zinc oxide-eugenol cement (ZEC) is used as the luting material, this risk appears to be lower, as the excess cement will dissolve in the peri-implant sulcus. However, using a ZEC on a general basis conflicts with the manufacturers' classification of a ZEC as a temporary luting material. To evaluate the clinical safety of ZECs, the present study investigated whether more ZECretained reconstructions than screw-retained reconstructions loosen over time. Method and Materials: In a retrospective observational study, the frequency of FDP loosening was investigated. For this purpose, data from patients who had been fitted with FDPs on two or more than two posterior implants were investigated. The study compared screw-retained (n = 59) vs cement-retained (n = 40) FDPs on implants over an observation period of 3.5 years after reconstruction incorporation. All cement-retained FDPs had been cemented with a temporary ZEC. Results: The prevalence of reconstruction loosening was significantly lower for cement-retained FDPs when a ZEC (10%) was used than it was for screw-retained FDPs (29%); four screw-retained FDPs and one cementretained FDP developed major complications. The survival rate of the reconstructions within the observation period was 97% for screw-retained FDPs and 100% for cement-retained FDPs (not significant). Conclusion: The prevalence of cementretained FDP loosening was significantly lower than that of screw-retained FDPs. Cementing FDPs on implants using a temporary cement does not necessarily lead to a higher number of loosened reconstructions. Permanent cementation with ZEC appears to be justified. Palabras clave: cement-retained, fixed dental prostheses, loosening, screw-retained, TempBond