Tasks such as retrieval practice or self-explaining are expected to trigger generative learning activities necessary for deep understanding. The present study investigated the influence of asking students to either perform re-read, retrieval, or self-explanation activities on text comprehension. Students of two German high schools were randomly assigned to one of three groups when reading expository texts. Subsequently, participants completed a posttest including retention and inference questions. An initial evaluation of the posttest data revealed no significant differences between the groups. This finding raised the question to what extent students actually carried out the requested learning activities and whether the quality of the performed learning activities plays an important role. The results indicate that qualitatively different learning activities are associated to different posttest performance. We conclude that explicitly examining traces of students’ learning activities might be a useful supplement in further research on generative learning activities.