In this paper we study how the rate of childcare utilization among families with children less than three years old is explained by differences in neighborhood income dispersion across given levels of household income. We form a pool of cross sectional databases using census block-group level averages and counts from the 1990 and 2000 decennial censuses and from the 2006-2010 American Community Survey. We calculated various measures of income dispersion among synthetic neighborhoods, which were created on the basis of geographical proximity between census block groups. For each census block-group in the nation, and anchoring on its geographic centroid, we calculate Euclidean distances to identify every block-group that surrounds the anchored block groups in a two miles radius. Our measures of neighborhood income dispersion reflect inequality in household income across block groups in close proximity. For each anchored block-group, we calculated the ratio of children 3 years old and older enrolled in preschool/nursery to the total number of children less than six years old (called the childcare utilization ratio). We also identified other relevant average demographic characteristics of the anchored block-groups, such as household type, educational attainment of adults, employment and poverty. Our empirical analysis cross tabulates the childcare utilization ratio by percentiles of both neighborhood income dispersion and average block group household income. We describe the conditional (using demographic characteristics) and unconditional distributions of the childcare utilization ratio for block groups with similar average household income across different levels of neighborhood income dispersion. Our findings help us to understand the role played by geographically determined spillover effects on utilization of childcare, across different levels of income. Patterns of utilization of childcare programs among similar low income households differ on the basis of income dispersion in their neighborhood. Conclusions and policy implications from our findings are discussed.