Learner corpora are electronic collections of authentic texts compiled from speech or writing by language learners that can be analysed automatically with the help of special software. While learner corpora, so far, have impacted the field of language pedagogy mainly in the fields of materials design (e.g. through informing the compilers of learner dictionaries or learner grammars) and in language testing and assessment (e.g. through informing current descriptor scales, such as the Common European Framework of Reference), the uptake of learner corpora in classroom practice, including for data-driven learning (DDL), is marginal compared to the use of native corpora. Therefore, in this entry, I will give an overview of how and why it can be beneficial to use learner corpora in DDL scenarios in the classroom.