This in vitro study evaluated the fracture resistance of teeth restored with different designs of partial ceramic restorations using two diameters of steel ball to apply fracture stresses. One hundred and twenty sound maxillary premolars were randomly divided into three groups of 40 elements; each group was submitted to one of three indirect restoration designs: inlay, onlay with only lingual cuspal coverage and onlay with buccal and palatal cuspal coverage. Another 20 intact teeth were randomly assigned as control groups. The restorations were produced with Super Porcelain EX-3 and Vitadur Alpha ceramics and luted according to manufacturers' instructions. The specimens were subjected to compressive axial loading in a universal testing machine at 0.5 mm/minute using two steel balls (3 and 10 mm in diameter), evaluating a total of 14 groups with 10 specimens each. Peak load to fracture was measured for each specimen. The results were submitted to analysis of variance and Tukey's test. Statistical analysis revealed that the inlays showed a significantly higher fracture resistance than both onlay designs, but with fracture resistance lower than that of intact teeth. Onlay fracture strength was equivalent for both designs. The force required to cause fracture with the 10-mm diameter ball was greater than with the 3-mm diameter ball. There were no differences between the tested ceramics.