This study examines the statistical association of foreclosure sales with social, economic and housing variables measured at the Census tract level for two purposes of interest to foreclosure mitigation practitioners —- to assess whether it is feasible to identify in advance neighborhoods likely to have high rates of foreclosure, and to explore the socioeconomic traits of high-foreclosure neighborhoods so as to design appropriate mitigation programs. We collected data on foreclosure sales in 2002 from the sheriff’s departments of Hennepin and Ramsey counties, the two core counties that comprise the Minneapolis-St. Paul MSA. We find that several factors commonly associated with high foreclosure sale rates could have correctly identified, in advance, most neighborhoods with high rates of mortgage foreclosure. To guide the design of foreclosure mitigation programs, we also present evidence that foreclosure risks in our two counties were highest in neighborhoods with elevated credit risk indicators and a high proportion of homeowners who are recent minority buyers or young. We show that an accurate credit risk variable is among the best predictors of foreclosure and also critically affects our multivariate analysis of factors associated with foreclosure. To limit social losses associated with foreclosures, we conclude that consideration should be given to enhancing public access to data on mortgages, foreclosures, and foreclosure risk factors, especially the neighborhood distribution of credit scores.