In this retrospective study, the authors investigated pediatric blood lead levels (BLLs) at 2 threshold levels in neighborhoods across the US city of Chicago, examining geographic associations with demographic risk factors and housing characteristics, using data from large-scale childhood BLL screening records from 1997 through 2003. They used logistic regression and geostatistical methods to assess disease dynamics and probability of elevated BLLs. The results showed a significant decline of elevated BLLs, with levels measured at >or= 10 microg/dL decreasing by 74%, compared with a 40% decrease for the lower levels (6-9 microg/dL). The Westside and Southside neighborhoods, with a high concentration of minority populations, had the highest prevalence rates, which were significantly associated with living in pre-1950 housing units. The findings provided insights for lead prevention, implications for lowering the threshold limit, and suggestions for the urgent task of developing healthy neighborhoods.