I examine the student-achievement implications of school-level budgetary autonomy in public schools. I propose a theoretical model of a principal's allocation decisions and empirically examine determinants of participation and the resulting impacts on student achievement. A principal's years of administrative experience positively impact program participation both for middle schools and elementary schools. In elementary schools, estimates show that increased autonomy improves both math and reading scores. Autonomous elementary schools with a high percentage of free and reduced lunch eligible students also appear to have a greater ability to improve test scores.