The purpose of this study was to evaluate the in vitro shear bond strengths to enamel and the failure sites of three ceramic brackets and one metal bracket in combination with light cured orthodontic adhesive. The brackets were divided into four groups. Each ceramic bracket group had different bonding mechanisms with adhesive. Group A; metal bracket with foil-mesh base (control group) Group B; ceramic bracket with micromechanical retention Group C; ceramic bracket with chemical bonding Group D; ceramic bracket with mechanical retention and chemical bonding. Forty extracted human lower first premolars were prepared for bonding and 10 brackets for each group were bonded to prepared enamel surfaces with Transbond^® light cured orthodontic adhesive. Twenty four hours after bonding, the Instron universal testing machine was used to test the shear bond strength of brackets to enamel. After debonding, brackets and enamel surfaces were examined under stereoscopic microscope to determine the failure sites. Statistical analysis of the data was carried out with ANOVA test and Scheffé test using SPSS PC+. The results were as follows: 1. There were statistically significant differences in mean shear bond strengths of three ceramic bracket groups (p<0.05). Shear bond strengths of group C and D were significantly higher than that of group B and shear bond strength of group C was significantly higher than that of group D. 2. Group C and D both had significantly higher shear bond strengths than metal bracket (group A), but there were no significant differences in shear bond strengths between group A and B (p<0.05). 3. The failure sites of four bracket groups were also different. Group C and D failed primarily at enamel-adhesive interface, but group A and B failed primarily at bracket base-adhesive interface. 4. Among all ceramic bracket groups, group B was very similar to metal bracket in the aspect of shear bond strength and failure site.