THE SUBJECTS WERE 'OFF-THE-STREET' ARRESTEES IN JAILS IN CHICAGO, ILL.; NEW ORLEANS, LA.; SAN ANTONIO, TEX.; BROOKLYN, N.Y.; LOS ANGELES, CALIF.; AND ST. LOUIS, MO. PERSONS ARRESTED FOR RELATIVELY MINOR OFFENSES (E.G., TRAFFIC VIOLATIONS, DRUNK DRIVING) WERE EXCLUDED FROM THE SAMPLE. DRUG USE DATA WERE OBTAINED FROM SELF-REPORTS, URINE ANALYSES, AND OFFICIAL RECORDS. MULTIVARIATE ANALYSIS OF VARIANCE TECHNIQUES ARE USED TO ANALYZE CRIME RATE PATTERNS WITHIN COHORTS OF DRUG USERS (I.E., THOSE ARRESTEES WHO BEGAN USING DRUGS DURING THE SAME TIME PERIOD) AND TO ANALYZE DIFFERENCES IN CRIME RATE PATTERNS ACROSS COHORTS. OVERALL ARREST RATES AND INVOLVEMENT IN SERIOUS CRIMES AGAINST THE PERSON ARE COMPARED FOR DRUG USERS AND NONUSERS. DRUG USERS HAVE A HIGHER OVERALL ARREST RATE THAN DO NONUSERS. SINCE 1967, THERE APPEARS TO HAVE BEEN A SHIFT IN THE TYPES OF CRIMES FOR WHICH DRUG USERS ARE ARRESTED AWAY FROM SERIOUS CRIMES AGAINST THE PERSON AND TOWARD PROPERTY AND OTHER FORMS OF CRIME. THE RESULTS ALSO POINT UP THE SIGNIFICANCE OF HOW THE CRIME OF ROBBERY IS CLASSIFIED. SUPPORTING DATA AND A LIST OF REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)