We compare introductory microeconomics courses that employ a blended learning approach to more traditional control courses at a selective liberal arts college. The blended learning courses required students to complete online homework and watch video lectures before problem-based class sessions, while the control courses used paper homework and primarily traditional lecturing. We use the Test of Understanding in College Economics to measure improvement and use the Student Opinion Survey to control for students' self-reported effort and amount of importance placed on the test. We find that students in the blended courses improve more than those in traditional courses.