Students typically create concept maps while they view the material they are trying to learn. In these circumstances, concept mapping serves as an elaborative study activity--students are not required to retrieve the material they are learning. In 2 experiments, we examined the effectiveness of concept mapping when it is used as a retrieval practice activity. In Experiment 1, students read educational texts and practiced retrieval either by writing down as many ideas as they could recall in paragraph format or by creating a concept map (retrieval-based concept mapping). In Experiment 2, we factorially crossed the format of the activity (paragraph vs. concept map) and the presence or absence of the text (i.e., whether the activity involved repeated studying or retrieval practice). On a final test 1 week later that assessed verbatim knowledge and inferencing, both paragraph and concept map retrieval practice formats produced better performance than additional studying, but the 2 retrieval formats themselves did not differ. The results demonstrate the effectiveness of concept mapping when it is used as a retrieval practice activity and show that retrieval itself, rather than merely the act of writing, drives the benefits of retrieval-based learning activities.