Purpose: The purpose of this paper is to measure the effects of COVID-19 on student learning. Using boundary theory and border theory, the authors hypothesize that learning will be highest in 2019, the year before the pandemic, and lowest in 2020, the year that the pandemic began. Design/methodology/approach: Participants were business students at an American university. The authors administered the same 88-question multiple-choice exam in a capstone course in May 2019, May 2020, and May 2021. Ten questions measured learning in the capstone course (current learning), and the remaining questions assessed learning from prerequisite courses (prior learning). Any year-to-year differences in test scores may be attributed to the effects of COVID-19. Findings: Current learning was exactly as hypothesized. It was highest in 2019 and lowest in 2020. Prior learning appeared to be completely unaffected by the pandemic. It increased slightly but insignificantly throughout the three years. The authors conclude that home-school boundary and border problems caused by the pandemic inhibited the ability of students to remember what they had recently learned but did not prevent them from demonstrating knowledge of information that they had acquired in pervious semesters. Originality/value: The authors add value to the body of knowledge about the effects of COVID-19 on student learning because (1) our dependent variable is cognitive. Most other studies have modeled attitudinal dependent variables such as satisfaction; and (2) this sample is located within the United States. Most other research has utilized participants in other countries and their results may not generalize to the American context.