This paper reports on a controlled experiment on the effects of three types of reflection triggers in an online course. Fifty-four volunteers, distributed in five groups, used these structured opportunities for reflection during learning. Results show that reflection triggers were extensively employed by the test persons and were perceived as quite useful to reflection and learning. Test persons in the experimental groups reported significantly more reflective prompting and more intensive reflection than those in the control group. In contrast, no positive effects on learner performance and retention could be established. This paradox elicits different possible explanations, which are discussed in the light of the common pedagogical claim that more thoughtful approaches to learning should be promoted.