We experimentally test the impact of student loan “nudges” on community college students’ borrowing decisions and subsequent educational attainment. We find that students are biased towards borrowing the amount listed in their financial aid award letters, even though this amount does not affect students’ choice sets. Students randomly assigned to receive a nonzero loan offer were 40 percent more likely to borrow than those who received a $0 loan offer. Neither fall nor spring enrolment was affected by loan offers, but students induced to borrow by a nonzero offer earned significantly more credits and higher GPAs. An additional $1000 in loans led to 0.9 additional credits earned and a 0.16 GPA increase in the first year. Given that nearly one quarter of U.S. college students are offered $0 in loan aid, our results indicate the potential to achieve large gains in educational attainment by reforming the choice architecture around borrowing