Abstract: The present study provided partial validation for Hartmann's (1991) boundary construct and its theorized relationship to dream content. Additionally, this study examined the complex relationship between presleep stimuli, stress, and boundary structure and what influence they have upon dream content. Participants were exposed to bizarre, affectively charged, and dramatic presleep video content for a period of 28 days; their dream reports were compared to a group of participants exposed to a set of control videos. The Boundary Questionnaire was used to measure boundary structure and the State-Trait Anxiety Inventory was used to measure stress. Composite dream content variables were measured by the Hall and Van de Castle System of Content Analysis. Subjective and objective dream report elements comprised the Emotions Composite and Dramatic Intensity Composite. These multi-method instruments were created by combining Hall and Van de Castle variables with global ratings scales to create two comprehensive dream measures. People with thinner boundaries reported increased levels of state-anxiety and increased levels of trait-anxiety. Interaction effects were observed between experimental condition, boundary thinness, and dream content. People who were experiencing increased levels of stress produced positive correlations between boundary thinness and dream negative emotions. Males reported more trait-anxiety than females. Females produced dream reports with more emotions and negative emotions than males. Results were examined within the context of continuity and compensatory dream theories. The adaptive functions of dreams were also considered.