The study defined and tested a model, based on pertinent Cooperative Extension Service literature, that was meant to project changes in present educational audiences well enough to predict educational needs, and would identify new audiences in the early stages and derive implications for the institution and its programs. Ninety-seven farm operators in Suffolk County, New York, were interviewed. Major findings included the following: (1) general population trends were not good indicators of definite educational programs; (2) population changes in clientele groups gave data on which to base educational programs; (3) in some situations, the predicted elimination or reduction of a group posed the immediate problem of shifting resources to other clientele groups; (4) a related problem was continuation of educational activities of acceptable quality to a dwindling audience; (5) forecasts made in testing the model tended to underestimate the actual change rate, but they were close enough to permit program planning for a group of reasonable size. (author/ly)