FICTITIOUS APPLICATIONS WERE IDENTIFIED AS YUGOSLAVIAN OR AUSTRALIAN NATIONALS WITH OR WITHOUT A NONSPECIFIED CRIMINAL BACKGROUND. STANDARD APPLICATION LETTERS WERE FORMULATED IN WHICH ONLY THE STUDY VARIABLES, JOB POSITION, DATE, AND SOURCE OF THE JOB ADVERTISEMENT WERE VARIED. ALL APPLICATIONS WERE POSTED ON THE MONDAY FOLLOWING THE SATURDAY ON WHICH THE ADVERTISEMENT APPEARED. THEY WERE SENT OVER A PERIOD OF 3 CONSECUTIVE WEEKS. A RECORD WAS KEPT OF THE TIME TAKEN FOR RESPECTIVE EMPLOYERS TO REPLY AND THE NATURE OF THE EMPLOYERS' RESPONSES. ONLY REPLIES RECEIVED WITHIN A 31-DAY PERIOD FROM POSTING WERE ASSESSED. DATA WERE ANALYZED BY RODGER'S (1969) GENERALIZED-MEDIAN PROCEDURE FOR PLANNED CONTRASTS. THERE WERE NO SIGNIFICANT DIFFERENCES EITHER IN THE MEASURES OF RESPONSE FOR MIGRANT (YUGOSLAVIAN) AND NONMIGRANT (AUSTRALIAN) GROUPS OR BETWEEN CRIMINAL AND NONCRIMINAL GROUPS. THE ONLY SIGNIFICANT ETHNIC/CRIMINAL INTERACTION EFFECT WAS FOUND WITH RESPECT TO THE PROPORTION OF NONREPLIES: PUBLIC AND PRIVATE COMPANIES REPLIED LESS OFTEN TO JOB APPLICATIONS FROM NONMIGRANT CRIMINALS THAN TO THOSE FROM MIGRANT CRIMINALS. THUS, THERE APPEARS TO BE A SLIGHT BIAS FAVORING MIGRANT EX-OFFENDERS OVER NONMIGRANT EX-OFFENDERS IN TERMS OF COURTESY REPLIES, ALTHOUGH NOT IN TERMS OF EMPLOYMENT PROSPECTS. POSSIBLE FACTORS WHICH MAY HAVE CONTRIBUTED TO THE LACK OF DIFFERENCE IN EMPLOYMENT PROSPECTS FOR CRIMINAL AND NONCRIMINAL GROUPS INCLUDE THE LITERACY AND WORDING OF CRIMINAL-GROUP APPLICATIONS. A SHORT PRISON SENTENCE WAS INDICATED, BUT THE LETTER WAS FRAMED IN A RESTITUTIONAL ATTITUDE: ('THE JOB IS VERY IMPORTANT TO ME ... AND (I) WOULD APPRECIATE THE CHANCE TO START WORK AGAIN'). TABLES CONTAINING RESPONSE AND ANALYSIS MATERIAL AND REFERENCES ARE INCLUDED. (JAP)