The aim of this preliminary in vitro study was to compare the retentiveness of a luting agent designed for use with dental implants to luting agents designed for use with tooth-retained restorations. The following luting agents were tested: (1) implant cement, (2) resin-bonded zinc oxide-eugenol cement, (3) zinc phosphate cement, (4) zinc polycarboxylate cement, and (5) glass-ionomer cement. After cementation, each sample was subjected to a pull-out test using a universal testing machine, and the loads required to remove the crowns were recorded. The mean values and standard deviations of cement failure loads were analyzed using analysis of variance and the Bonferroni test. The mean cement failure loads (N) were 333.86 ± 18.91 for implant cement, 394.62 ± 9.76 for resin-bonded zinc oxide-eugenol cement, 629.30 ± 20.65 for zinc phosphate cement, 810.08 ± 11.52 for zinc polycarboxylate cement, and 750.17 ± 13.78 for glass-ionomer cement. The retention provided by polycarboxylate cement was significantly greater than that of all other luting agents; the implant cement showed the lowest retention values. These preliminary in vitro observations need to be confirmed under conditions that more closely approximate the clinical environment.