THE PURPOSE OF THIS ANALYSIS WAS TO IDENTIFY THE DEFENDANTS MOST LIKELY TO RECIDIVATE WHEN THEY FIRST ENTER THE CRIMINAL JUSTICE SYSTEM AND TO EVALUATE DECISIONS MADE BY THE POLICE, PROSECUTOR, DEFENSE COUNSEL AND JUDGE DURING THE PROCESSING OF A CASE, IN TERMS OF THEIR EFFECTS ON THE PROBABILITY, FREQUENCY, AND SERIOUSNESS OF RECIDIVISM. FIVE MEASURES OF RECIDIVISM WERE USED. DATA WAS BASED ON THE 15,460 CASES BROUGHT TO THE PROSECUTOR IN 1973, INVOLVING 12,382 DEFENDANTS. ONE OF THE PRIMARY FINDINGS OF THE ANALYSIS WAS THAT PERSONAL CHARACTERISTICS OF THE DEFENDANT AND CHARACTERISTICS OF HIS CRIMINAL HISTORY WERE MORE IMPORTANT DETERMINANTS OF RECIDIVISM THAN ACTIONS TAKEN BY THE CRIMINAL JUSTICE SYSTEM DURING CASE PROCESSING. SUCH CHARACTERISTICS AS RACE, AGE, SEX, EMPLOYMENT STATUS, ALCOHOL ABUSE, AND OPIATE USE WERE SIGNIFICANTLY RELATED TO FREQUENCY AND SERIOUSNESS OF FUTURE CRIMINAL ACTS. THE ITEMS IN THE DEFENDANT'S CRIMINAL HISTORY WERE ALL STATISTICALLY SIGNIFICANT. THE CRIMINAL PROCESSING VARIABLES WHICH SEEMED TO AFFECT RECIDIVISM ARE ALSO DISCUSSED.