YEARS. IT IS NOTED THAT THE CRUCIAL ISSUE RELATING TO THE USE OF OFFICIAL RECORDS IN DEVIANCE RESEARCH IS WHETHER BIASING FACTORS ARE RANDOM (TENDING TO CANCEL OUT EACH OTHER) OR SYSTEMATIC (REDUCING VALIDITY FOR RESEARCH TO UNACCEPTABLE DEGREES). FROM A REVIEW OF LITERATURE ON HOW SPECIFIC VARIABLES INFLUENCE THE ACCUMULATION OF DIFFERENT TYPES OF DEVIANCE STATISTICS, FOUR GENERAL SOCIOLOGICAL FACTORS ARE INDICATED TO EMERGE AS SOURCES OF BIAS IN SUCH INDICES: (1) THE NATURE OF PARTICULAR DEVIANT ACTIVITIES; (2) THE CLIMATE OF SOCIAL CONTROL; (3) THE SOCIAL ORGANIZATION OF AGENCIES PRODUCING THE INDICES; AND (4) THE INTERACTIVE PROCESSES BETWEEN THESE AGENCIES, THE PUBLIC, AND POTENTIAL DEVIANTS. AFTER DISCUSSING THESE FACTORS FROM A GENERAL PERSPECTIVE, A STUDY OF THEIR INFLUENCE ON DRUG ARREST STATISTICS COMPILED BY THE NARCOTICS UNIT OF THE CHICAGO POLICE DEPARTMENT OVER A 30-YEAR PERIOD (1942-1970) IS DESCRIBED AND DISCUSSED. A 25 PERCENT SAMPLING RATIO WAS USED FOR THE YEARS 1942-1949 (TO INCREASE N'S), WHILE A 12.5 PERCENT RATIO WAS USED FOR THE LAST 2 DECADES WHEN THE TOTAL NUMBER OF ARRESTS WAS LARGE. AN OVERALL SAMPLE OF MORE THAN 18,000 (OF AN ESTIMATED 140,000) DRUG-RELATED ARRESTS WAS OBTAINED. SUCH INFORMATION AS PLACE OF ARREST, TYPE OF CHARGE, OFFENDER'S RESIDENCE, AGE, SEX, AND RACE WAS INCLUDED IN THE RECORDS. THE PLACE OF ARREST WAS CLASSIFIED BY CENSUS TRACT, AND ARRESTS WERE CLASSIFIED ACCORDING TO THE RACE OF THE OFFENDER. THE ARREST RATE PER 100,000 OF THE APPROPRIATE BASE POPULATION WAS CALCULATED FOR EACH CENSUS TRACT. THESE RATES PROVIDED A STANDARDIZED INDEX FOR COMPARISONS OF ARREST PATTERNS. SIGNIFICANT CHANGES IN PATTERNS WERE ANALYZED BY A STUDY OF POLICE RECORDS, INTERVIEWS WITH RETIRED NARCOTICS OFFICERS, AND OBSERVATION OF POLICIES AND PROCEDURES OF CURRENT NARCOTICS ENFORCEMENT. ON THE BASIS OF THE STUDY, IT IS ARGUED THAT THE STATISTICS COMPILED REFLECT SYSTEMATIC BIASES IN THE OPERATIONS OF POLICE ASSIGNED TO THE NARCOTICS DIVISION. IT IS CONCLUDED, THEREFORE, THAT RESEARCHERS OUGHT NOT TO RELY ON OFFICIAL ARREST RECORDS AS INDICES OF COMPARATIVE DRUG ACTIVITY OVER A LONG PERIOD OF TIME. (RCB)