Abstract. Due to the complexity of economic system, the interactive effects of economic variables or factors on Chinese foreign trade make the prediction of China's foreign trade extremely difficult. To analyze the relationship between economic variables and foreign trade, this study proposes a novel nonlinear ensemble learning methodology hybridizing nonlinear econometric model and artificial neural networks (ANN) for Chinese foreign trade prediction. In this proposed learning approach, an important econometrical model, the cointegration-based error correction vector auto-regression (EC-VAR) model is first used to capture the impacts of the economic variables on Chinese foreign trade from a multivariate analysis perspective. Then an ANN-based EC-VAR model is used to capture the nonlinear patterns hidden between foreign trade and economic factors. Subsequently, for introducing the effects of irregular events on foreign trade, the text mining and expert's judgmental adjustments are also incorporated into the nonlinear ANN-based EC-VAR model. Finally, all economic variables, the outputs of linear and nonlinear EC-VAR models and judgmental adjustment model are used as another neural network inputs for ensemble prediction purpose. For illustration, the proposed ensemble learning methodology integrating econometric techniques and artificial intelligence (AI) methods is applied to Chinese export trade prediction problem.