The analysis used data for a probability sample of 1,888 cases from approximately 100,000 begun by arrests on felony charges, covering every major crime category in New York City's four major boroughs in 1971. Additional information came from interviews with police officers, prosecutors, defense attorneys, and judges involved in additional probability sample of 369 felony arrests reaching dispositions in 1973 to determine the reasons for dispositions. Findings revealed a system dominated by plea bargaining, in which only 2.6 percent of the cases ended in trial. Unexpectedly, the charge reductions and dismissals were often explicable, according to participants, by factors other than caseloads and related pressures. However, dispositions were generally proportional to offense seriousness, the defendants' criminal records, and the closeness of their relationships to the victims. Thus, the system was proportional in its outcome, although not in its process. Results also revealed that criminal conduct often results from problems in personal relationships among neighbors, friends, and former spouses. The reluctance of the complainants in these cases to pursue prosecution accounted for a larger proportion of the high rate of dismissal than any other factor. Findings suggested the desirability of making explicit the terms of the agreement reached by plea bargaining and the reasons for a sentence. Additional findings and implications, figures, tables, and footnotes