THE ADDICTS WERE INTERVIEWED TO OBTAIN QUANTITATIVE DATA ON NARCOTIC USE, EMPLOYMENT, AND ILLICIT BEHAVIOR DURING THE PERIOD FROM 12 MONTHS PRIOR TO FIRST NARCOTIC USE TO THE TIME OF THE INTERVIEW. ARREST DATA WERE COLLECTED FROM STATE RECORDS FOR THE SAME PERIOD. THE ARREST DATA AND SELF-REPORTED CRIMINAL BEHAVIORS WERE THEN EXAMINED TO DETERMINE THE FREQUENCY OF DRUG USE DURING ADDICTION CAREERS (FIRST DAILY NARCOTIC USE TO LAST DAILY USE). DATA COMPARISONS INDICATED THAT 35 PERCENT OF THE ADDICTION CAREER NONINCARCERATED TIME INVOLVED LESS-THAN-DAILY OR NO NARCOTIC USE. DURING THESE PERIODS, ARREST RATES FOR PROPERTY CRIME AND SELF-REPORTED CRIMINAL BEHAVIOR WERE SUBSTANTIALLY LOWER THAN FOR PERIODS OF DAILY USE. AN IMPORTANT ASPECT OF THESE FINDINGS IS THAT THEY ARGUE AGAINST THE OFTEN EXPRESSED VIEW THAT CRIMINALITY CONTINUES AS A LEARNED BEHAVIOR FOLLOWING THE CESSATION OF ADDICTION. WHILE THIS SEEMS TO BE TRUE TO SOME EXTENT, THE RESULTS INDICATE THAT AT LEAST THE AMOUNT OF CRIMINALITY TENDS TO BE SHARPLY REDUCED FOLLOWING SUCCESSFUL TREATMENT. IT IS ALSO NOTED THAT THIS APPROACH TO EXAMINING THE RELATIONSHIP BETWEEN CRIME AND NARCOTIC ADDICTIONS AVOIDS THE USUAL METHODOLOGICAL PROBLEMS ENCOUNTERED IN PREADDICTION AND POSTADDICTION AND PRETREATMENT AND POSTTREATMENT STUDIES. TABULAR DATA AND REFERENCES ARE PROVIDED. (KBL)