Background: Nonsignificant findings are reported in approximately half of the studies examining correlations between environmental characteristics and walking. One reason could involve spatial incongruence in which environments examined are not necessarily the ones where the behavior occurred. Purpose: To study correlates of perceived neighborhood size and the congruence between perceived and investigator defined neighborhood areas. Methods: Door-to-door surveys were conducted in 12 U.S. Census block groups with 18- to 87-year-old adults. Participants were asked about exercise participation and to draw a map representing their neighborhood. Geographical information software was used to construct maps, calculate neighborhood areas, and analyze congruence. Results: Neighborhood sizes perceived by women were smaller than those perceived by men and positively related to min/wk of biking and education. Education and age were negatively and perceived neighborhood support for exercise was positively associated with perceived neighborhood size in men. Approximately 90% of perceived neighborhood areas overlapped with neighborhood areas defined by circles with 1200-m radii from participants' homes. Conclusions: Several individual characteristics including exercise are associated with perceptions of neighborhood size. Standard investigator-defined neighborhoods using a 1200-m radius (15 min walk) from a participant's home is sufficient to capture a majority of perceived neighborhood areas.