For the problem of group multi-attribute decision making, in which the attribute values given by the decision makers are the intuitionistic fuzzy sets, a new method of determining the experts’ weights is proposed. Instead of the traditional experts’ weights measure, which is characterized by using the discriminating power to calculate the experts’ weights, the proposed method adopts the intuitionistic fuzzy entropy measure to emphasize the fuzzy degree of the information given by decision makers. Based on the four entropy measures, which are hesitation degree, distance, similarity and uncertainty degree entropy, the effects of experts’ weights results are compared in a computational experiment. The experiment results show that the experts’ weights not only depend on the different entropy measures but also the number of experts and attributes.