The crime rates examined include robbery, all violent crimes, breaking and entering, all property crimes, all criminal code offenses, and total offenses. Controls for age, gender, unemployment rate, and the consumer price index are included as well as a time trend variable. Both lagged and unlagged correlations are examined. Findings suggest that increased levels of per capita police strength precede increased crime rates, rather than being a response to increased crime rates. A 310-percent increase in police reported rates of violent crime between 1962 and 1988, however, represents an exception to this generalization. Lower levels of per capita police strength are associated with higher subsequent rates of violent crime. Results support the interpretation that increased levels of violence provide the rationale for increased levels of per capita police strength, but that these increased levels of police strength merely generate increased police activity which has not necessarily been well-targeted at containing violent crime. Various hypotheses and explanations of the findings are evaluated in terms of the effectiveness of community-based intervention to reduce violence. Supplemental data on crime rates and police strength are appended. 78 references, 37 endnotes, and 10 tables