For multi-attribute decision making problems, a new evaluation method for alternative selection using information axiom is proposed, of which the functional requirements are incomplete or unknown. Through identifying the type of each evaluation index and determining the probability of success satisfying the functional requirement of each index, the total information content for each alternative is obtained, and ranked in order, so the optimal alternative can be selected. For requirement of evaluation, the calculation formula of information content is amended. Further, this approach allows the decision-maker to amend design range according to the customer’s needs, so the satisfactory alternative with designer’s preference can be obtained and the evaluation process is placed into a more flexible and natural framework. Finally, an example is used to illustrate the proposed method.