THE QUESTIONS OF CENTRAL INTEREST TO THE RESEARCH ARE AS FOLLOWS: WHAT ARE THE CRITERIA FOR ELIMINATING OFFENSES AND OFFENDERS FROM THE PROSECUTION PROCESS? AND, WHICH OF THE PENAL SOCIAL-CONTROL INSTITUTIONS DECIDES WHICH PERSONS WILL BE SELECTED FOR PROSECUTION? DATA FOR THE STUDY WERE OBTAINED BY MEANS OF THREE EMPIRIC METHODS CONSIDERED TO COMPLEMENT AND CONTROL EACH OTHER -- ANALYSIS OF A SAMPLE OF 4,588 DOSSIERS REFERRING TO CRIMES AGAINST PROPERTY AND OTHER ASSETS, CARRIED OUT IN 8 PROVINCIAL COURT JURISDICTIONS OF THE FEDERAL REPUBLIC OF GERMANY; NONSTANDARDIZED INDIVIDUAL INTERVIEWS WITH 79 POLICE OFFICERS AT THE LOCATIONS OF THE REGIONAL COURTS; AND SUBSEQUENTLY, 8 GROUP DISCUSSIONS WITH THE INTERVIEWEES WERE HELD. THREE FEATURES OF THE OFFENSE WERE FOUND TO BE OF PARTICULAR IMPORTANCE IN CLEARING IT -- THE VISIBILITY OF AN OFFENSE TO THE VICTIM OR THE POLICE; THE PROBABILITY OF SOLVING A CASE; AND THE VARYING EXTENT OF EVIDENCE PROBLEMS. THESE JURIDICIAL PROBLEMS WERE FOUND TO OCCUR MOST FREQUENTLY IN CASES OF FRAUD AND EMBEZZLEMENT, AND OFTEN END WITH THE QUASHING OF THE PROCEEDINGS. SELECTION DETERMINED BY CERTAIN SOCIAL FEATURES OF A SUSPECT WAS FOUND TO BE RARE. IT WAS FOUND THAT THE POLICE, RATHER THAN THE PUBLIC PROSECUTOR, CONTROLLED THE PRELIMINARY PROCEEDINGS. RARELY DO THE PUBLIC PROSECUTORS' OFFICES INTERVENE IN POLICE INQUIRIES, WHICH ARE AUTONOMOUSLY CONDUCTED BY THE POLICE FORCES. THE OUTCOMES OF INVESTIGATIONS, THEIR SUCCESS RATE, THEIR FAILURES, MISTAKES, AND OMISSIONS HAVE A NOTICEABLE AND DECISIVE INFLUENCE ON PROSECUTION PRACTICES. (RCB)