Although a number of studies have analyzed ADR programs in the court and general contexts, very little quantitative analysis of police mediation system design exists, particularly regarding the types of complaints that are considered appropriate for mediation. The academic literature on police mediation has merely attempted to summarize the personal opinions of experienced mediators, taken somewhat extreme theoretical positions, or rehashed this data while trying to fit it into generalized mediation theory. As civilian agencies for processing civilian complaints against police officers have been in existence for, in some cases, over 25 years, this may seem surprising. However, many civilian oversight agencies did not initially have a formal mediation program (most still do not), and the number of complaints referred to mediation, even in the oldest programs, remained low until the last five to ten years. Thus, the amount of data available on mediation has probably not been sufficient, until recently, to support a statistically significant analysis of the relationship between the fundamental nature of a complaint and an eventual successful outcome in mediation, however one chooses to define that success. Though this article does not perform such an analysis, it will attempt to offer more detailed hypotheses regarding the types of complaints that are appropriate for mediation in the context of law enforcement. The bases of these hypotheses in many cases are data collected in studies of community or court mediation programs. Most existing data deals with non-criminal matters, but several studies have examined mediation in the context of criminal allegations by a complainant.