The fantasies, 'explanations', and drawings of seven patients' versions of their illnesses (four cases of diabetes, one cystic fibrosis with pneumonia, one laceration of an arm, and one severe asthma) comprise the data in this study. All patients experienced the illness as an insult to or imperfection in a previously intact body image. In every instance actual body changes were trivial in comparison to those depicted by the patients. This was most striking in diabetic patients whose illness affords no 'perceptual localization'. After a discussion of the evolution of body image, the author describes the complex character of its graphic representation. Children's reality knowledge and prior information about anatomy and illness are used in their drawings much as day residue is used in dreams. The further back one goes developmentally, the less probable it is that visual memories correspond accurately with original bodily experience. Kaufman warns against the common tendency toward direct correlating of graphic representations and actual experiences, or direct correlation of drawings of body