To pronounce words correctly is one of the most important elements in language learning. Children can learn pronunciation more effectively when teachers provide sufficient modelling and correct their erroneous utterances properly. If children’s erroneous pronunciations are left untreated, the errors will fossilize and be difficult to correct. Despite its vital role, research on how teachers treat young learners’ pronunciation errors is still scant. Therefore, this study aims to investigate what corrective feedback teachers use to treat young learners’ pronunciation errors and the reasons behind the decision. Case study is employed as the research method with two teachers differing in educational background and teaching experiences selected as participants. Lyster and Ranta’s (1997) categorization of corrective feedback was employed in this research. The findings demonstrate that recast and explicit correction are predominantly used to treat young learners’ erroneous pronunciation. The action mainly is unintended, but the secondary consideration is due to the characteristic pronunciation learning which, unlike grammar learning, is considered to be void of systematic thinking process. This study is expected to enrich the theories of corrective feedback and pronunciation teaching as well as to help teachers to decide what types of corrective feedback appropriate to treat their students’ pronunciation errors.