The purpose of this study was to evaluate the implementation of conferencing as a restorative policing practice. Family group conferencing is considered an important new development in restorative justice practice as a means of dealing more effectively with young first-time offenders by diverting them from court and involving their extended families and victims in conferences to address their wrongdoing. Cases deemed eligible for the study were property crimes including retail and other thefts, criminal mischief and trespass, and violent crimes including threats, harassment, disorderly conduct, and simple assaults. A total of 140 property crime cases and 75 violent crime cases were selected for the experiment, with two-thirds of each type randomly assigned to a diversionary conference (treatment group) and one-third of each type assigned to formal adjudication (control group). Participation in the conference was voluntary. If either party declined or if the offender did not admit responsibility for the offense, the case was processed through normal criminal justice channels. Those cases constituted a second treatment group (decline group). The Bethlehem, Pennsylvania, Police Department and the Community Service Foundation conducted a two-year study on the effectiveness of police-based family group conferencing. Beginning on November 1, 1995, 64 conferences were conducted for the study. Approximately two weeks after their cases were disposed, victims, offenders, and offenders' parents in the three experimental groups (control, conference, decline) were surveyed by mail, in-person interviews, or telephone interviews. Those who participated in conferences (Parts 4, 6, and 8) received a different questionnaire than those whose cases went through formal adjudication (Parts 5, 7, and 9), with similar questions to allow for comparison and some questions particular to the type of processing used on their case. Disposition data on cases were collected from five district magistrates in Bethlehem from January 1, 1993, to September 12, 1997. Data on recidivism and outcomes of the control and decline group cases were obtained from (1) the Bethlehem Police Department arrest database (Part 1) and (2) a database of records from the five district magistrates serving Bethlehem, drawn from a statewide magistrate court database compiled by the Administrative Office of Pennsylvania Courts (Part 2). An attitudinal and work environment survey was administered to the Bethlehem Police Department on two occasions, just before the conferencing program commenced (pre-test) and eighteen months later (post-test) (Part 3). Part 1 variables include offender age, year of offense, charge code, amounts of fine and payments, crime type, offender crime category, and disposition. Part 2 collected disposition data on cases in the study and officers' observations on the conferences. Demographic variables include offender's age at current arrest, ethnicity, and gender. Other variables include type of charge, arrest, disposition, sentence, and recidivism, reason not conferenced, current recorded charge class, amounts of total fines, hours of community service, and conditions of sentence. Part 3 collected information on police attitudes and work environment before and after the conferencing program. Variables on organizational issues include ratings on communication, morale, co-workers, supervision, administration, amenities, equipment, and promotions. Variables on operational issues include ratings on danger, victims, frustration, external activities, complaints, workload, and driving. In Parts 4 to 9, researchers asked offenders, parents of offenders, and victims about their perceptions of how their cases were handled by the justice system and the fairness of the process, their attitudes and beliefs about the justice system, and their attitudes toward the victim and offender. Variables include whether the respondent was satisfied with the way the justice system handled the case, if the offender was held accountable for the offense, if meeting with the victim was helpful, if the respondent was surprised by anything in the conference, if the respondent told the victim/offender how he/she felt, if there was an opportunity to reach an agreement acceptable to all, if the offender/parents apologized, if the victim/parents had a better opinion of the offender after the conference, what the respondent's attitude toward the conference was, if the respondent would recommend a conference to others, if the offender was pressured to do all the talking, if the offender was treated with respect, if victim participation was insincere, if the respondent had a better understanding of how the victim was affected, if the victim only wanted to be paid back, and if conferences were responsive to needs.