The study examined the delivery of police services in Rochester, N.Y.; St. Louis, Mo.; Tampa, Fla.; and St. Petersburg, Fla. Results show only a slight tendency for police to respond with greater speed to serious calls. Response to in-progress calls, for instance, was typically more rapid than response to calls not in progress, but not by very much. Few of the differences between types of calls achieve statistical significance. Response to a call of any type was typically remarkably swift. A somewhat more pronounced pattern of prioritized response was evident in the manner in which officers drive to the scene. About half of all in-progress calls prompt faster than normal speed in three of the cities. Still, this differentiation in response produces little variation in response time. This is attributable, at least in part, to the spatial concentration of patrol officers. A correlation of neighborhood service response times with socioeconomic characteristics, situation variables, and officer attitudes, indicates that the only significant nonrandom determinant of response time is the seriousness of the reported problem. This suggests that differential response times are prompted by professional norms rather than officers' personal biases. Tabular data, 8 footnotes, and 23 references are provided.