Decades of public policy have reduced airborne lead exposure but overlook water-based exposure, particularly in schools. Even low levels of lead harm children’s development, yet effects of school-age exposure remain unknown. This paper evaluates Baltimore City Public Schools’ 2007 switch from tap to bottled drinking water over lead contamination concerns. Using stacked synthetic control methods, I find this policy increased testing proficiency by 4-6 percentage points, with larger effects for lower-achieving, Title I eligible, and less diverse schools. These findings highlight the educational consequences of in-school environmental hazards, underscoring the need to address aging school infrastructures.