This paper studies the impact of fasting on students’ learning outcomes in the Indonesian context, the world's largest Muslim country. Precisely, we measure the effects of Ramadan fasting by using a simple 2 × 2 difference-in-differences setting, exploiting the fact that, in 2018 and 2019, the Ramadan period coincided with the final exam periods in Universitas Indonesia. Our study uses Universitas Indonesia's comprehensive student database (SIAK-NG), which provides detailed test scores, student-level, and course-level characteristics, and details on test schedules. It allows us to investigate Ramadan fasting's direct effects, differential effects (whether morning and afternoon test effects vary or not), and cumulative effects (whether or not effects vary over time). We find no evidence of Ramadan's negative effects on students’ test scores after controlling for semester-course-class fixed effects, student, class, and course characteristics. Consistently, we also find no evidence of morning-afternoon differential effects. We argue that the absence of (or weak) Ramadan's effects is likely because Muslim students had adapted to the fasting environment earlier in the Ramadan period, allowing them to improve their scores as the exam period progresses. The regression results also suggest that the fasting effects are more positive among students with less good academic performance. © 2022 Elsevier B.V.