Between late 1995 and late 1997, the British Home Office Police Research Group was working with Leicestershire Constabulary on a demonstration project that implemented problem-oriented policing in one of its areas. Repeat calls for service furnished an important starting point in police-led problem identification and analysis. The Police Research Group used four contrasting localities to examine their patterns of repeat calls for service. The localities included a rural area, a market town, a problem inner-city housing estate, and a mixed residential and shopping center. The analysis of repeat calls for service distinguished between crime-related calls; those arising from community/disorder problems; and a generic third category that covered social issues, alarms, and traffic. It also distinguished between patterns of repeat calls from residential addresses and nonresidential addresses. Nonresidential addresses consistently had much higher rates of repeat calls than residential addresses. The findings suggest that there are consistent variations in repeat-call patterns that reflect differences in routine activities according to land usage. These in turn yield contrasting problem types for police services that use problem-oriented policing. Police analysis, in which only rudimentary automated data-cleaning has been undertaken, has produced many informative findings relevant to problem-oriented policing. This suggests there is scope for practical, routine police use of incident data. 7 tables, 3 figures, and 22 references