Health communicators are often charged with the difficult task of communicating complex medical or scientific information to lay audiences. In the case of Phase III cancer clinical trials, communication efforts must help audiences understand both the medical and scientific aspects of treatment. Previous research has found that aversion to the scientific component, specifically randomization to a treatment condition, is a primary reason why individuals choose not to enroll in clinical trials. As a result, healthcare providers often use metaphorical language as a linguistic strategy to increase understanding and hopefully acceptance of randomization to treatment among patients. However, the effectiveness of this message strategy to produce outcomes favorable to clinical trial enrollment is unknown. Guided by McGuire's (1972) model of information processing, this study examined outcomes associated with randomization messages including attention, comprehension, yielding, arousal, and behavioral intention among a medically underserved audience of rural, low-income, older adult women. Formative research conducted with the intended audience led to the development of stimulus videos about randomization. A four-message condition between subjects pretest-posttest laboratory study was used to empirically test the effectiveness of these messages. Culturally tailored metaphor messages were compared to a conventional metaphor, a scientific definition, or attention control on outcomes specified in McGuire's model for information processing. No main effects for message strategy were found, however, message strategy significantly interacted with attention to predict posttest behavioral intentions. Implications of these findings for rural health message design are discussed.