Interventions included interagency training and information classes, investigative checklists for police, a legal manual for police, prosecutor charging and disposition standards, police prescreening of cases, quarterly feedback reports, and followup letters and personal visits by the prosecutor to the police. A comparison of pre- and postintervention data for 1984-1986 indicate that during the followup period, the Prosecutor's Office filed felony charges for only 51 percent of referrals from law enforcement agencies. The remaining referrals were either declined entirely or filed as misdemeanors. The interventions had no effect on either avoidable or unavoidable attrition. An unexpectedly low rate of declines was attributed to avoidable attrition: only 15 percent of attrition was judged to be avoidable and potentially correctable. An analysis of system, victim/witness, evidence, and defendant-related reasons for attrition indicates that system factors accounted for a majority of declines, followed by victim/witness, and defendant reasons. Attrition rates, both avoidable and unavoidable, varied among types of felony offenses, with the majority of avoidable attrition occurring in the least serious felony offenses. 6 figures, 11 tables, and 11 references.