This paper reports and discusses the results of a study aimed at automatically categorising teacher feedback on student writing. A total of 3412 teachers’ written comments on 90 students’ draft essays were collected from an EFL course offered by a Hong Kong university during the first semester of 2016/17. The data were primarily used to design and implement an automated tool to classify teachers’ comments with respect to a taxonomy of their characteristics. The findings of this study show that the performance of the automated tool is comparable to that of human annotators, suggesting the feasibility of using the automatic approach to identify and analyse different types of teacher feedback. This study can contribute to future research into the investigation of the impact of teacher feedback on student writing in a big data world. Copyright © 2018 International Journal of Information and Education Technology. All rights reserved.