Abstract: This study investigated a hypothesis that preference for sensory modality of imagery, also referred to as "system representation", is an important dimension of adequate communication between a therapist and the client, one which would be expected to influence the outcome of therapeutic procedures. The subjects for this experiment were 40 Emergency Service Operators, civilian employees of the Detroit Police Department. These operators were believed to experience high levels of stress due to the nature of their occupation, responding to telephone requests for emergency police, fire, and ambulance assistance. The subjects were 38 females and 4 males, ranging in age from 21 to 64, with one to four years of experience as an Emergency Service Operator. All had at least a high school education. Subjects were administered two measures developed for this study: the Auditory-Visual Imagery Inventory and the Emergency Service Operator Stress Scale; a standardized measure of psychiatric symptomatology, the System Check List-90; and a standardized measure of job satisfaction, the Minnesota Satisfaction Questionnaire. Subjects were divided into those who preferred visual imagery and those who preferred auditory imagery. Experimental subjects were assigned to one of four groups, reflecting a combination of subject preference for imagery and type of treatment: auditory preference - visual imagery; auditory preference - auditory imagery; visual preference -auditory imagery; and visual preference -visual imagery. There was also an untreated control group. All 32 treated subjects received one relaxation training and four systematic desensitization sessions, worded either to emphasize auditory or visual imagery. All measures were then readministered. There was no evidence to support the hypothesis that subjects treated by a method worded to be congruent with their preferred system of representation would show greater treatment effects than subjects treated by a method worded to be incongruent. This was interpreted as casting doubt on the suggestion that a client's system of representation is an important dimension in communication between therapist and client. The systematic desensitization technique had some positive effects on reducing job related stress. There was an interesting finding that those subjects who preferred auditory imagery reported fewer and less severe symptoms. This was interpreted as suggesting that preference for imagery may be a factor in adjusting to occupational demands. Further research on this possibility is indicated.