Lead paint is a pervasive and harmful environmental hazard, yet childhood lead poisoning incidence has been declining. Due to high clean-up costs, lead-poisoning prevention programs in the United States rely on self-selection into health screenings to identify sources of lead exposure. Is self-selection into prevention programs an effective targeting mechanism? I study screening take-up by analyzing geocoded 2001-2016 data on 2 million Illinois children, 2.9 million blood lead tests, and 4,500 healthcare providers. Using fine-grained geographic fixed effects, I find that children living farther from providers, especially high-quality providers, are less likely to be screened, and those who do get tested do not have higher blood lead levels. I derive households’ willingness-to-pay (WTP) for screening from travel costs and perform counterfactual policy simulations. Average WTP among high-risk households appears low, at $6.14, consistent with low baseline poisoning rates. Despite poor targeting, screening incentives to households or providers may improve welfare because of the large externalities of lead exposure.