Based on the relationship between the marginal distribution function（MDF） and the joint distribution function（JDF） of the two-dimensional random variable（TRV）, this paper concerns how to build the JDF of multidimensional random variables （MRV） copula function, and for the construction of function fitting and inspection. First, it introduces the definition of Copula function, the types of SKlar theorem and commonly used functions of the Copula Function. Then the relationship between the MDF function and JDF function of TRV variables is extended to MRV variables. Finally this paper studies the building JDF-function method under the MDF-function already of the TRV- variables, and the method extend to the MRV variables. It also puts forward the new method of constructing JDF function of MRV variables, and the corresponding methods of fitting and parameter estimation of the function constructed.