The problem of forecasting monthly demands for library network services is considered in terms of using forecasts as inputs to policy analysis models, and in terms of using forecasts to aid in the making of budgeting and staffing decisions. Box-Jenkins time-series methodology, adaptive filtering, and regression approaches are examined and compared within a real context of forecasting demands for interlibrary loan services within the Illinois Library and Information Network (ILLINET). The evaluation of each forecasting method utilizes data for total ILLINET interlibrary loan demands. Developed as a result of the method comparisons, a set of interactive computer programs are listed which will allow the application of forecasting methods by a person who is not an expert in forecasting methodologies or in use of the computer. These programs include data entry, parameter estimation, and production of forecasting reports, and the design of each is briefly summarized. It is concluded that regression methods are most accurate and easiest to apply, and that improvements in forecasting accuracy can be achieved by using a fading memory method. Tables, equations, and a list of references are provided. (FM)