We adapted and evaluated the effects of a classroom intervention (extension of https://bibliotheek.ehb.be:2102/10.1016/j.learninstruc.2018.04.006) on undergraduates' sourcing skills. Students (n = 266) received either a teacher-led intervention (trained group) or regular classes (control group) and were assessed before, after, and 6 to 8 weeks after the intervention. Additionally, we included an ecological indicator: the quality of references in a high-stakes team essay. Overall, the intervention group outperformed controls in both post-tests: They gave lower ratings for bad quality links and used source-based criteria more often to select a text deemed as more reliable when reading two conflicting documents. They also included better quality references in their essays (i.e. references applying the targeted source dimensions) and avoided lesser quality ones (i.e. references overlooking those dimensions). No differences were found in the mention of sources in a far transfer argumentation. We discuss that the intervention improved students' sourcing skills even in long-term generalized tasks.