Statement of problem Clinical studies have shown that lack of retention is one of the major causes of fixed dental prosthesis failure. Purpose The purpose of this study was to determine the relationship between the convergence angle of a complete metal crown tooth preparation and the recementation strength for restorations cemented with conventional and adhesive cements. Material and methods One hundred twenty artificial crowns were cast for standardized complete metal crown tooth preparations accomplished with the use of a milling machine on extracted human teeth. Three different tapers, 5, 12, and 25 degrees, were used (n=40). The crowns in each group were subdivided into 4 subgroups (n=10) according to the luting cement: zinc phosphate cement (Fleck's), glass ionomer cement (Ketac-Cem), and adhesive resin cements (Panavia 21 and C&B-Metabond). Retention was evaluated by measuring the tensile force required to dislodge the crowns from the tooth preparations in a universal testing machine. Subsequently, the tooth preparations were scraped clean and polished with prophylaxis paste, and the intaglio surfaces of the artificial crowns were ultrasonically cleaned and airborne-particle abraded with 50-μm aluminum oxide powder prior to recementation. The data were analyzed with 3-way repeated-measures ANOVA and Tukey's HSD test (α=.05). Results Analysis of measurements disclosed a significant difference for taper and luting cement ( P<.001); however, their interaction was not significant. Also, there was a statistically significant difference between the retention of the first cementation and the second cementation ( P<.001). However, the interaction was not significant with taper or cement. Regardless of the taper used, Panavia 21 cement exhibited the highest mean initial retention, but the difference was not significantly different from the recementation retentive strength. Conclusions Tooth preparation taper and type of luting cement had a direct effect on the recementation strength of complete metal crowns.