Exposure to lead through drinking water is of concern for children, particularly at schools and early education and care facilities (EECFs), where they spend much of their time. We use lead and copper data from monitoring in schools and EECFs in Massachusetts (USA) and create risk indices based on the percentage of fixtures in a school above three water lead level (WLL) thresholds (15, 5, and 1 ppb) to model which building characteristics, water source, and water treatment practices are associated with a school exceeding these thresholds. Local building characteristics had larger effects than information about the public water supplier (PWS), and buildings built from 1950 to 1980 were most at risk. Daily flushing and fixture replacement often decreased elevated WLLs, and water coolers had lower WLLs than other fixtures. These findings highlight the value of WLL monitoring programs and can be used to prioritize future investment in monitoring and remediation. (author abstract)