Using data from the 2000 U.S. Census, members of the Wichita Police Department and the researcher developed beat profiles for each of the city's 36 patrol beats. Currently, these profiles estimate the population, proportional racial and ethnic group representation, and per capita income of each patrol beat. The data showed a racially segregated pattern of residences within the city, with most beat populations being primarily non-Black and a few beats (four) composed primarily of Black residents. During the first 6 months of 2001, the department collected information on 37,454 police stops. The analysis of the data found moderate to high correlations between the race of individuals stopped by police and the predominant racial composition of the beat. Officer deployment was found to be influenced by the crime rate and population of the beats; the predominate race of the beat did not apparently influence the level of officer deployment. In beats predominately populated by non-Black residents, stops of Black citizens were overrepresented (proportionally); and in beats predominately populated by Black residents, stops that involved non-Black citizens were also overrepresented (proportionally). Based on the findings of this study, a theory of "contextual attentiveness" is developed to explain the racial pattern of police stops. This theory proposes that police officers use the circumstances associated with a distinct episode or location to define what is usual, customary, or expected within that particular context. Officers then become differentially attentive to individuals or behaviors that appear to them to be inconsistent with predetermined concepts of what is usual, customary, or expected within a particular context; therefore, when police officers observe individuals in their beat area whose race is different from the predominate racial composition of the community, whether Black or non-Black, the chance that these individuals will be stopped by officers increases. 4 tables and 35 references