Students and instructors show moderate levels of agreement about the quality of day-to-day teaching. In the present study, we replicated and extended this finding by asking how correspondence between student and instructor ratings is moderated by time of semester and student demographic variables. Participants included 137 students and 5 instructors. On 10 separate days, students and instructors rated teaching effectiveness and challenge level of the material. Multilevel modeling indicated that student and instructor ratings of teaching effectiveness converged overall, but more advanced students and Caucasian students converged more closely with instructors. Student and instructor ratings of challenge converged early but diverged later in the semester. These results extend our knowledge about the connection between student and faculty judgments of teaching.