Objective: Two different mechanisms involving resin cements have been proposed as possible ways of strengthening all-ceramic crowns. These are (1) modification of the intaglio surface by crack healing and (2) strengthening of the ceramic by shrinkage of the resin cement, which would stress the molecules together after polymerization. This in vitro study tested the validity of these two proposed mechanisms. Materials and Methods: Two hundred 12-mm-diameter by 2mm-thick Vitadur Alpha disks (VITA)-an aluminous oxide layering porcelain-were fabricated. The disks were ground with silicon carbide abrasive papers and then allocated into groups of 20 specimens as follows: 1. No treatment was performed. Specimens were stored in a desiccator to serve as dry controls. 2. Specimens were indented to produce a 30- to 40- mu m indentation. As in the first group, specimens were then stored in a desiccator to serve as dry controls. 3. No treatment was performed, and specimens were stored in a water bath at 37 degrees C for 24 hours to serve as wet controls. 4. Specimens were stored in a water bath at 37 degrees C for 24 hours to serve as wet controls. In contrast to the previous group, these specimens were indented. 5. Ground specimens were etched with Mirage Super Etch (Chameleon Dental Products Inc., Kansas City, KS, USA)-a 9.6\% hydrofluoric acid gel-for 90 seconds. Monobond-S silane agent (Ivoclar Vivadent AG, Schaan, Liechtenstein) was applied to the acid-etched surface for 60 seconds and air-dried. The treated surface was then coated with Compolute Aplicap (3M ESPE)-a dual-cured resin cement-following the manufacturer's instruction. The cement was light-activated with an Optilux, 501 unit (Demetron Kerr). 6. Specimens were treated as in the preceding group, but also were indented. 7. Specimens were treated as in the Compolute group except that RelyX Unicem Aplicap (3M ESPE) was the resin cement used. 8. RelyX Unicem Aplicap was the resin cement used, and specimens were indented. 9. Indented specimens were etched only to determine the effect of etching. 10. Indented specimens were etched and primed to determine the effect of these procedures. All specimens except for those in the first two groups were kept in a water bath at 37 degrees C for 24 hours until testing. Blaxial flexure strength testing was performed in all specimens. Profilometry evaluation was done to examine the surface roughness of the ground and indented control specimens. Results: Both Compolute and RelyX Unicem significantly increased the fracture strength of ground (20 and 42\%, respectively) and indented specimens (28 and 48\%, respectively) when compared with the respective wet controls. Dry controls had higher fracture strength than their respective wet controls. Etching and priming did not have any effect on fracture strength of the indented specimens. Conclusions: The authors concluded that the strengthening of the porcelain occurs by shrinkage of the resin cement, and does not depend on the severity of defects.