Since the introduction of ceramic brackets to orthodontic therapy, a need has arisen to test the manufacturer's claims regarding these brackets. Forty-eight noncarious human canine teeth were divided equally into groups A to D. Brackets were bonded to these teeth with the use of the acid-etch technique and a composite resin according to the manufacturer's instructions. The combination within each group was as follows: A = stainless steel brackets and chemically cured resin; B = ceramic brackets and chemically cured resin; C = ceramic brackets and light-cured resin; D = stainless steel brackets and light-cured resin (via transillumination). After curing, the teeth were stored for 1 week in distilled water at 37 degrees C. The Instron machine was used to test the shear bond strengths of the brackets to the teeth. The brackets were individually tested to failure of the bond, which was recorded along with the site of fracture. The conclusions are as follows: (1) all combinations produced shear bond strengths that were greater than those that are considered clinically acceptable, (2) the ceramic groups exhibited a significantly higher bond strength than that of the stainless steel group, and (3) enamel fractures occurred among the B group in 40% of the samples tested in that group. It is thus apparent that a fracture of enamel is a real possibility during therapy or at debonding of the ceramic brackets, especially if the tooth is nonvital.