Abstract: Effects of a sensory based rapport method and chemical dependency workshop on knowledge and sentiments of adults employed in Kansas social service agencies were studied. The rapport method experimental group experimental group would not have significantly higher mean scores than controls, on measures of chemical dependency course knowledge, sentiment toward the experimental trainer, or sentiments about the workshop's usefulness; that they would not rate the experimental method trainer significantly higher than the seven control trainers; that all subjects would not score significantly higher on the knowledge posttest, than on the pretest, and that age, sex, job category, and education would have no main interactive effects on the three criterion measures (knowledge, usefulness of workshop, or sentiment toward trainer). Data was pooled from four deliveries of the course in Topeka and Kansas City during 1982. An accessible sample of 79 subjects (59 women and 20 men) employed in Kansas Service Agencies of Topeka, and Kansas City, participated in 50 hours of course instruction over a five week period. The participants were randomly assigned to either experimental or control conditions. Forty experimentals and 39 controls completed the workshop, from an original sample of 42 experimentals and 45 controls, resulting in an attrition rate of controls that was three times higher than that of the rapport group experimentals. Analysis of variance revealed significantly higher mean scores for the rapport group over controls on measures of workshop usefulness and sentiment toward the experimental trainer, and no significant difference on the course knowledge post test. Significant main effects of education and age were noted on the course knowledge post test. Although no demographic variables were interactive with the experimental rapport method on the course knowledge post test or usefulness of the workshop measure, three variables (education, sex, and job category) were interactive with the rapport method on sentiment toward the experimental trainer. A t-test for related samples indicated a significantly higher (p<.0005) course knowledge post test mean over the pre test mean for all subjects. An effect size (.0287) of superiority of the rapport method over the control (traditional) instruction method was noted. Subjects rated the experimental rapport method trainer significantly higher, on an analysis of variance, than all seven control trainers (p<.001).