The maintained hypotheses embodied in structural general equilibrium models calibrated to data have tended to make economists and policy makers insecure regarding their empirical foundation. Advances in dynamic general equilibrium (DGE) theory and its empirical application have exacerbated this insecurity since the forecasts provide by these models brings questions of validation to the forefront. Here, methods are developed to measure the magnitude of bias in DGE forecasts that are simple to implement. We adopted the concordance correlation measure, and introduced a time function method to assess the bias in DGE forecasts. A time-series confidence interval method is also introduced to formally judge the "good" forecasts from the "bad". A calibrated DGE model is used to illustrate them. The time function method allows for the choosing of a functional form and an upper bound on forecast error. The time-series confidence interval method allows the DGE results to be evaluated by the standard of the rival time series models. If the DGE results are as good as time-series forecasts, the DGE model is a superior framework because of its advantage in providing not only "good" forecasts, but also insights into the economic structure generating the results. To illustrate these methods, we calibrate to Taiwanese data for the year 1988 a multi-sector Ramsey-based DGE model. The model is shown to forecast various dimension of the economy with surprising good, but varying accuracy. The proposed validation measures show effectiveness in distinguishing among diverse model parameter values and detecting model improvements. The measures are also statistically meaningful and require no arbitrary probabilistic assumptions on the distribution of either the results or the data.