In a large introductory physics course, structured weekly journals (weekly reports) regularly encouraged students to ask questions about the material. The resulting questions were collected for one quarter and coded based on difficulty and topic. Students also took several conceptual tests during the quarter. The reports contained more questions than typically observed in a college classroom, but the number of questions asked was not correlated to any measure of conceptual performance. Relationships among different types of questions and performance on these tests were explored. Deeper-level questions that focus on concepts, coherence of knowledge, and limitations were related to the variance in student conceptual achievement. © 2003 Wiley Periodicals, Inc. J Res Sci Teach 40: 776–791, 2003