Due to rising costs and declining affordability, many students have to work while attending college. The federal government takes a major role in subsidizing the wages of college students and spent over $1 billion on the Work-Study program in 2010-11 (College Board, 2011), yet little is known about how working during the school year impacts college student outcomes. Few studies have explored the causal effect of the Federal Work-Study program, and the previous studies provide an incomplete picture. In contrast to these earlier studies, our paper focuses on a large sample of over 45,000 students in the Ohio public higher education system, which is the fifth largest public system in the United States and reflects the national mixture of selective and nonselective four-year institutions. We estimate the causal effects of participating in the Federal Work-Study program on a range of college students' academic outcomes, including college GPA, credits earned, and persistence. Our empirical strategy employs a differences-in-differences instrumental variables model, which exploits the variation in allocation of federal work-study funds across institutions and across students. We find that working on-campus has a small, negative, but statistically significant effect on students' first-year GPAs. However, we also find some evidence that participating in the Federal Work-Study program increases the number of credits that students accumulate by the end of their first year. These results suggest that working in an on-campus, work-study job may have small negative effects that are outweighed by potential benefits. One figure and 7 tables are appended.