The purpose of this study was to investigate whether teachers' ability to identify bullying incidents could be enhanced by offering bullying identification trainings. The participants of this study were 532 primary, middle, and high school teachers in Taiwan, who were recruited to participate in one of the four study groups (three treatment groups and one control group). Those in the three treatment groups were provided with the following interventions: a 1.5-h training on bullying identification (Group X); a definition including the features of bullying in written form (Group Y); and a written definition accompanied by checklist of three bullying characteristics for each scenario (Group Z). A 24-item Recognition of Bullying Incidents Questionnaire (RBIQ) was used in this study. Of these 24 scenarios, 12 were used to assess participants in the pre-test phase, and 12 were used for the post-test. The results, using a mixed-model two-way analysis of variance, indicated that the teachers in Group X were significantly better at bullying identification than those in other groups. Teachers in Group Z also were significantly better at bullying identification than teachers in Group Y and the control group. These results imply that merely providing a definition does little to enhance teachers' ability to identify school bullying.