Abstract: In the light of the uncertain factors in the investment decision system, a model for multiple attribute decision analysis (MADA) based on evidential reasoning (ER) is developed to solve the problems with both quantitative and qualitative information in uncertain environment. With the ER, the confidence matrix is used to express the assessed values of attributes, and a rule based on such a technique is developed to transform various types of information according to the grade of the general attribute so that all the attributes can be assessed in a consistent manner. The ranking of scheme is finally generated on the basis of the magnitude of utilities by virtue of utility theory. The investment decision is exemplified to demonstrate that the ER is effective in investment decision system.