Business Improvement Districts (BIDs) provide supplemental services to urban commercial corridors using funds from member assessments. They have become a very popular urban revitalization tool, but their formation is still largely unexplained. Theory implies that BIDs will form if they add to aggregate welfare and if the marginal net benefit of membership is positive. I test this for the neighborhood overall and at the BID boundary. Using unique, micro-level and longitudinal data from New York City, I employ survival analysis methods to estimate the likelihood of a neighborhood forming a BID. I then estimate the likelihood of the marginal property’s BID membership by comparing the characteristics of properties located immediately inside and outside of the BID boundaries. I find that BIDs are more likely to form when there is more commercial space over which the BID benefits can be capitalized and when there is homogeneity in service and spending preferences across properties. BIDs also tend to form in neighbor- hoods that possess signs of appreciation and growth. Generally, BIDs are more likely to form in neighbor- hoods with higher valued properties with the exception of very wealthy areas. The BID boundary, however, is comprised of relatively less valuable properties.