Question generation is theorized to support comprehension, self-regulation, and achievement, yet the empirical base for whether and how student-generated questions are associated with comprehension monitoring and whether they predict performance remain open questions. To address these, we investigated the questions undergraduate students in an introductory chemistry course recorded in question logs across an 8-lecture unit and their relations with post-lecture self-appraisals of comprehension and exam performance. Results indicated that students who generated more questions during lectures, who were able to resolve fewer of their questions, and who generated questions indicating large exam-relevant knowledge gaps reported lower levels of comprehension after lectures. Questions that sought verification or disambiguation were positively associated with exam performance. Findings suggest that student-generated questions can be meaningful indicators of comprehension monitoring processes and academic performance.