Price indices related to specific expenditure categories are necessary for realistic budget projections in higher education. Given the erratic inflationary behavior of the past seven years, and the uniqueness of higher education expenditures, realistic inflation prediction requires a balanced combination of analytic forecasting and intuitive projection. Methodology developed for inflation prediction at the University of Massachusetts/Amherst prepares price indices tailored to the University budget. Forecasts are based on analysis of inflation trends since 197l, and allow for subjective weighting of trends by a budget analyst. The data base used for analytic forecasting is a set of quarterly price indices selected from readily available sources. University price indices are derived that correspond to specific expenditure categories. Long-term, medium-term, and short-term inflation predictors are analytically developed. The budget analyst supplies subjective weighting to the three inflation predictors to obtain an overall inflation predictor for each expenditure category. The paper includes a description of the analytic methodology and discussion of implementation in the budget planning process. (Author)