Abstract: The accuracy of a single forecasting model is not high, and the forecasting results can not reflect the relationship between economic variables. For solving these problems, the gray multivariable linear econometric model is constructed by combining with the gray GM(1,1) model and multivariable econometric model. And then, the proposed model is used to forecast trend of the per capital net income of rural residents in Hunan province. The forecasting precision of the proposed model is not only higher than the single gray GM(1,1) model and multivariable linear regression model, but also could reflect the relationships between the economic variables. The forecasting results show that the per capital net income of rural residents in Hunan province will increase year by year in next few years, but the growth rate will be relatively slow. Finally, some suggestions for regulating the income gap of rural residents in Hunan province.