The author uses regression and matching techniques to evaluate Indonesia's Social Safety Net Scholarships Program, which was developed to keep large numbers of children from dropping out of school as a result of the Asian crisis. It was expected that many families would find it difficult to keep their children in school and that dropout rates would be high, as they were during a recession in the 1980s. But dropouts did not increase markedly and enrollment rates remained relatively steady. The author examines the role the scholarship program played in producing this result. She found the scholarships to have been effective in reducing dropouts in the lower secondary school (where students are more susceptible to dropping out) by about 3 percentage points. They had no discernible impact in primary and upper secondary schools. The author also examines how well the program adhered to its documented targeting design and how effective that design was in reaching the poor. Committees that allocated the scholarships followed the criteria diligently, but a significant percentage of scholarships did go to students from households with high reported per capita expenditures, if household expenditure data are reliable. It is unclear how targeting can be improved, giving the scarcity of accurate local household data in most countries. Using local monitoring could help but then monitoring for accountability would be more difficult. Preliminary evidence favors focusing safety net scholarships--designed to reduce dropout rates during an economic crisis--on lower secondary schools, continuing to target children (especially older students) from large families, scaling back scholarships to private schools at the lower secondary level, or targeting the households hurt most by the crisis.