Comprehension of science topics occurs when learners meaningfully generate relationships and conceptions about what they read. In this generation process, learners' cognitive and metacognitive regulation is one of the most critical factors influencing learning. However, learners are not always successful in regulating their own learning, especially in computer-based learning environments (CBLEs) where they are alone. Based on this rationale, the present study was designed to examine the effects of two scaffolding strategies--generative learning strategy prompts and metacognitive feedback--on learners' comprehension and self-regulation while learning the human heart system in a CBLE. Participants were 223 undergraduate student volunteers. Structural Equation Modeling (SEM) was employed to conceptualize and empirically test a model that explains mediating processes among variables. Results revealed that the combination of generative learning strategy prompts with metacognitive feedback improved learners' recall and comprehension by enhancing learners' self-regulation and better use of highlighting and summarizing as generative learning strategies. (Contains 6 tables and 6 figures.)