In order to deal with multiple experts’ judgment information, incomplete information and inconsistent information in C4ISR system efficiency evaluation, an approach is proposed for obtaining criteria weights under incomplete information environment. The approach is made up of five steps: collecting the multiple experts’ opinion and constructing judgment matrices; defining incomplete coefficient matrices; constructing multiple objectives optimization model by minimizing the deviation degrees; resolving model using goal programming method; and analyzing result. Finally, a numerical example for sub-criteria weighs determination is studied and shows the advantages that the proposed approach can integrate multiple experts’ information, avoid inconsistent and obtain most satisfied criteria weights successfully.