STATEMENT OF PROBLEM: Bond stability between zirconia crowns and luting cement and between cement and dentin is a main concern; however, only limited evidence is available as to its longevity. PURPOSE: The purpose of this in vitro study was to measure the retentive strengths of 7 self-adhesive cements (RelyX Unicem Aplicap, RelyX Unicem Clicker, RelyX Unicem 2 Automix, iCEM, Maxcem Elite, Bifix SE, SpeedCem), 2 adhesive cements with self-etch primers (Panavia 21, SEcure), 1 glass ionomer cement (Ketac Cem), 1 resin-modified glass ionomer cement (Meron Plus), and 1 zinc phosphate cement for luting zirconia crowns (LAVA) to extracted teeth after thermocycling with or without 1 year of water storage. MATERIAL AND METHODS: Two-hundred-forty extracted human molars (2 treatments; n=10 per cement) were prepared in a standardized manner. All cements were used according to the manufacturers' recommendations. The intaglios of the crowns were treated with airborne-particle abrasion. After thermocycling (×5000, 5°C/55°C) with or without 1 year of water storage, the cemented ceramic crowns were removed by using a Zwick universal testing device. Statistical analyses were done with the Wilcoxon rank sum and the 2-independent-samples Kolmogorov-Smirnov test. RESULTS: Median retentive strengths [MPa] for specimens thermocycled only/thermocycled with 1 year of water storage were as follows: Panavia 21: 1.7/2.5, SEcure: 3.0/3.0, RelyX Unicem Aplicap: 3.1/3.4, RelyX Unicem Clicker: 4.1/4.2, RelyX Unicem 2 Automix: 3.8/3.1, iCEM: 2.3/2.7, Maxcem Elite: 3.0/3.2, Bifix SE: 1.7/1.7, SpeedCem: 1.3/1.6, Meron Plus: 3.1/2.7, Ketac Cem: 1.4/1.4, and zinc phosphate cement: 1.1/1.6. Statistically significant differences were found only among specimens thermocycled only or thermocycled with 1-year water storage (P<.001). CONCLUSIONS: Significant differences in retentive strengths were observed among cements after thermocycling only or thermocycling with 1 year of water storage, but not for the effect of the additional 1 year of water storage.