THE SUBJECTS RECORDED THE CONTENT OF THEIR DREAMS DURING THREE PERIODS: (1) FOLLOWING THE INITIATION OF THE RELEASE PROCESS, (2) JUST BEFORE THE PAROLE BOARD APPEARANCE, AND (3) AFTER APPROVAL AND BEFORE RELEASE. THE GREATEST STRESS WAS EXPECTED TO OCCUR JUST BEFORE THE BOARD APPEARANCE. FURTHER, DREAM CONTENT STEMMING FROM ANXIETY, HOSTILITY, AND PERCEPTIONS OF BEING UNDER EXTERNAL CONTROL WERE EXPECTED TO VARY WITH DEGREES OF STRESS. DREAMS WERE ANALYZED FROM A JUNGIAN PERSPECTIVE FOR EVIDENCE OF COMPENSATION AND THE APPEARANCE OF TRANSITION SYMBOLISM. ELEMENTS OF ACTUAL EXPERIENCES WERE EXPECTED TO BE INCORPORATED IN THE DREAM CONTENT. FINDINGS CONFIRMED THE HYPOTHESIS THAT ANXIETY VARIED WITH THE STAGES IN THE RELEASE PROCESS. THE EXPECTED DREAM CONTENT RELATED TO ANXIETY, HOSTILITY, AND PERCEPTIONS OF EXTERNAL CONTROL WAS NOT CONFIRMED. MORE THAN HALF THE DREAMS WERE ABOUT RELATIONSHIPS, RELEASE, AND DRUGS, SUBJECTS WHICH WERE COMPENSATORY FOR CONSCIOUS ATTITUDES. AS HYPOTHESIZED, TRANSITION THEMES WERE FREQUENTLY IN THE SUBJECTS' DREAMS. THE EVIDENCE SHOWS THAT DREAM INTERPRETATIONS CAN BE A NONTHREATENING TECHNIQUE FOR BRINGING INMATES IN TOUCH WITH THEIR FEELINGS SO THAT THEY CAN MANAGE THEMSELVES MORE EFFECTIVELY. SUGGESTIONS FOR FURTHER RESEARCH ARE OFFERED. THE APPENDIXES INCLUDE SAMPLES OF THE INSTRUMENTS USED IN THE STUDY AND SUBJECTS' ACCOUNTS OF THEIR DREAMS. TABULAR DATA AND REFERENCES ARE PROVIDED.