This book presents the methodological approach to constructing econometric forecasting models for tax revenues. The modeling is carried out using the software EViews and is described in detail. In particular, it analyzes how well different models are suited for forecasting the revenue from the value-added tax (VAT) collected in Austria. The analysis is conducted within the framework of a model and method comparison and includes various forecasting horizons. The examined spectrum of econometric methods ranges from dynamic structural single-equation approaches of classical econometrics to model classes of univariate and multivariate time series analysis. The empirical results provide evidence that econometric modeling approaches are highly effective for forecasting VAT revenues. The mean absolute prediction error (MAPE) of the best econometric models reaches a value of 0.83% for 1-year forecasts and 0.77% for dynamic 3-year forecasts during the testing period. The advantages are also evident in the higher precision compared to the tax forecasts of the Federal Ministry of Finance (MAPE: 2.4%).