THE EVALUATION DESIGN CALLED FOR HEROIN ADDICTS ENTERING THE HOSPITAL'S DETOXIFICATION WARD OVER AN 18-MONTH PERIOD TO BE ASSIGNED RANDOMLY TO FOUR DRUG PROGRAMS. FOLLOWUP INTERVIEWS WERE TO BE CONDUCTED 6, 12, AND 24 MONTHS AFTER THE INITIAL INTERVIEW IN ORDER TO ASSESS HOW THE SUBJECTS WERE DOING IN TERMS OF DRUG USE, EMPLOYMENT, AND INCOME. TWO RESEARCHERS LOCATED AND COLLECTED DATA ON APPROXIMATELY 600 SUBJECTS. THE FIRST TWO FOLLOWUP CONTACTS WERE DONE PRIMARILY THROUGH THE MAIL AND OVER THE TELEPHONE. THE 24-MONTH CONTACT, FOR WHICH THE SUBJECTS WERE PAID, WAS DONE IN PERSON. WAYS IN WHICH THE RESEARCHERS USED THE INITIAL INTERVIEW TO ESTABLISH RAPPORT WITH THE SUBJECTS, TO IMPRESS UPON THEM THE IMPORTANCE OF THEIR PARTICIPATION, AND TO GATHER INFORMATION THAT WOULD ASSURE FOLLOWUP CONTACTS ARE DESCRIBED. PREPARATION OF FOLLOWUP QUESTIONNAIRES AND THE USE OF ROLE PLAYING TO PREPARE FOR TELEPHONE CONTACTS ARE DISCUSSED. SOURCES USED TO LOCATE SUBJECTS -PRISON RECORDS, MOTOR VEHICLE DEPARTMENT RECORDS, DIRECTORIES THAT LIST TELEPHONE NUMBERS BY STREET ADDRESS RATHER THAN NAME -- ARE NOTED, AS ARE METHODS USED TO MAKE CONTACT IN PARTICULARLY DIFFICULT CASES (CERTIFIED MAIL, MAILGRAMS, NIGHTLETTERS, 'BOMBARDMENT' WITH LETTERS AND TELEPHONE CALLS). REFERENCES ARE NOT INCLUDED. (LKM)