The author concludes that police-generated crime statistics may be an excellent tool for evidence-based policymaking. "Evidence-based practice" (EBP) involves making practice decisions based on the integration of the best research evidence available and applied under the expertise of practitioners with a view toward clients' attributes, values, preferences, and circumstances. EBP as a process enables political decisionmakers to focus on evidence-based cause- and- effect correlations based in research. In the first part of this paper, the theoretical bases and assumptions of knowledge-based policymaking are illustrated in discussing the aspects of police-generated crime statistics by using assumptions of scientific theory and method. From this perspective, statistical data depend on theory, because the arbitrary collection of data is not scientific. Data also have an empirical character, because they stem from real events. Further, data are objective, because collection processes must be factually and traceably explained. Official crime statistics measure only those crime cases and case features known to police. The unreported cases cannot be represented in the figures. Thus, police crime statistics are unreliable in identifying the causes of crime. Although police-generated crime statistics are currently unsuitable for more than purely descriptive conclusions, it is still possible, by monitoring them over a certain time period, to identify development trends that may vary significantly with different groups and areas. This enables a general presentation of the future tasks of the security forces and the government department in charge of public safety. A meaningful interpretation of police crime statistics in terms of evidence-based policymaking must collect information that results in identifying trends and also in clarifying the effectiveness of political governance. 34 notes and 23 references