Most past studies on the urban-rural differences in consumption structure implicitly assumed that the urban-rural differences are self-evident, and, as a result, these researches are not founded on sound statistical basis. In light of this negligence, by putting dummy variables into a basic AIDS model and testing the significance of their coefficients, a t Test is performed in order to give a statistical conclusion as to whether there exist significant differences in consumption structure between urban and rural residents, which would be useful as reference for related researches. Results show that significant differences only exist in some types of expenditures.