A REVIEW OF THE STUDIES ON POLICE, ATTORNEY, AND JUDICIAL BEHAVIOR AS TO WHETHER THEIR REACTIONS VARY WITH THE SOCIAL ATTRIBUTES OF THE ALLEGED DEVIANT OR CRIMINAL LEAD TO THE CONCLUSION THAT NO CONSENSUS EXISTS CONCERNING THE EFFECTS OF DECISIONS AND DISCRETIONS WHICH OCCUR PRIOR TO SENTENCING. THE DATA FOR THIS STUDY WERE COLLECTED FOR ALL CRIMINAL HOMICIDES WHICH OCCURRED IN A LARGE NORTHERN INDUSTRIAL CITY IN 1972. OF THE 383 CASES ANALYZED, 258 WERE CLASSIFIED AS SOCIAL CONFLICT HOMICIDES AND 125 AS CRIME-SPECIFIC HOMICIDES. THE DEPENDENT VARIABLE WAS THE DISPOSITION OF THE CASE AT THE PRELIMINARY HEARING; DEMOGRAPHIC VARIABLES INCLUDED OCCUPATIONAL STATUS, RACE, AGE, AND EDUCATION; CRIMINAL HISTORY VARIABLES WERE NUMBER OF FELONY PROPERTY CONVICTIONS, NUMBER OF FELONY PROPERTY ARRESTS, NUMBER OF AGGRAVATED ASSAULT CONVICTIONS, AND NUMBER OF AGGRAVATED ASSAULT ARRESTS. AFTER BIVARIATE ANALYSIS WAS PERFORMED AND THE MULTIPLE REGRESSION METHOD WAS USED, IT WAS FOUND THAT THE ATTRIBUTES OF THE PARTICIPANTS IN SOCIAL CONFLICT HOMICIDE ARE NOT AS POWERFUL IN PREDICTING DISPOSITION AS ARE THE ATTRIBUTES OF PERSONS INVOLVED IN CRIME-SPECIFIC HOMICIDE. THE OCCUPATION (OR LACK OF IT), AGE, THE CRIMINAL RECORD OF THE OFFENDER, AND OCCUPATIONAL STATUS OF THE VICTIM INFLUENCE POSTARREST REACTIONS OF CRIMINAL HOMICIDE OFFENDERS. IN PARTICULAR, ONLY ONE SOCIAL ATTRIBUTE, OCCUPATIONAL STATUS OF THE VICTIM, INFLUENCES LEGAL OUTCOMES. THE UNEMPLOYED OFFENDERS HAVE THEIR CASES PROSECUTED TO A SIGNIFICANTLY GREATER DEGREE THAN DO THE EMPLOYED OFFENDERS. CONVERSELY, WHEN THE VICTIM IS EMPLOYED, OFFENDERS HAVE THEIR CASES PROSECUTED TO A GREATER DEGREE THAN IF THE VICTIMS ARE UNEMPLOYED. NOTES, REFERENCES, AND TABULAR DATA ARE PROVIDED. (WJR)