This study investigates whether online homework improves learning in principles of microeconomics. It contributes to the related literature in two dimensions. First, it tests whether there is a positive association between students' performances on homework and exams. Importantly, it measures learning through exams that were given shortly after students completed the relevant homework, thereby avoiding a potential "decaying effect." Second, it tests whether the impact of homework varies by learning mode--face-to-face versus online. Results show that online homework improves learning. However, there is no significant difference in the impact of homework between the two learning modes.