This study evaluated the load bearing capacity of direct and indirect veneers vs. full-coverage crowns and classified the failure types after fracture load. Sound human maxillary incisors (N=108, n=12 per group) were randomly divided into nine groups to receive one of the following restoration types: Group 1: Intact tooth, Group 2: Direct resin composite, Group 3: Lingual: Indirect composite veneer, Labial: Ceramic veneer, Lingual overlap: Ceramic, Group 4: Lingual: Indirect composite, Labial: Ceramic, Lingual overlap: Indirect composite, Group 5: Lingual: Direct composite, Labial: Ceramic, Group 6: Lingual: Ceramic, Labial: Ceramic, Group 7: Feldspathic ceramic crown, Group 8: Metal-ceramic Crown, Group 9: Lithium disilicate crown. Teeth were prepared simulating the erosion/wear conditions in each group. After cementing, the specimens were stored in distilled water at 37 degrees C for 2months and then loaded to failure from the lingual surface at 105 degrees inclination in the Universal Testing Machine (1mm/min). Failure types were classified as irreparable or repairable. Data were analyzed using one-way ANOVA, Sheffe and Bonneferroni tests (=0.05). Mean fracture strength (N) of Groups 1, 4, 8, and 9 (558 +/- 278 - 880 +/- 319) were significantly higher than those of other groups (348 +/- 101-421 +/- 162) (p<0.05). Lingual veneering with direct/indirect resin composite or ceramic did not significantly affect the results (p>0.05) but lingual overlap with indirect composite increased the results (p<0.05). Group 1, 2, 4 and 5 presented more repairable failures. Restoration of eroded teeth could best be achieved with direct composite veneer at the lingual and ceramic veneer on the labial surface.