This study assesses the alternate-form reliability of the Body Type Dictionary (BTD) for measuring body boundary imagery and primordial thought language in interview-based and web-based Rorschach responses. The intraclass correlation coefficient demonstrated fair to good agreement for barrier imagery, .72, and penetration imagery, .55, which indicates that the web-based administration of the Rorschach inkblot test represents an acceptable alternative to the traditional Rorschach interview assessment for measuring body boundary imagery. Primordial thought language had a fair level of agreement, .43, whereas conceptual thought language had poor agreement, .36. The results are discussed by relating empirical research outlining the mode-specific implications of psychometric test administration to the Rorschach inkblot test and its implications for body boundary awareness and regressive cognitive functioning, as well as by outlining the methodological and clinical limitations of web-based Rorschach application that could be addressed in future research.