An interval evidential reasoning method was proposed to solve the problem of the evaluation level of attribute information and the uncertainty of evaluation level trust in the process of multiple attribute decision making. Firstly, the uncertainty of evaluation information was expressed in the form of interval evaluation level and interval evaluation level trust. Secondly, the trust recognition framework of attribute evaluation was unified by transforming formulas. Then, the integration formula of interval evidential reasoning was used to aggregate the interval uncertain information, to deal with the problem of the aggregation of the interval uncertainty information of the evaluation level, the evaluation level of trust and the quantitative attribute evaluation. Finally, the scheme was sorted synthetically by adopting the theory of expected utility. In order to verify the effectiveness of the proposed method, the proposed method was used to process the uncertain evaluation information in the form of interval, taking a fresh supplier selected by a fresh food E-commerce as an example. The result shows that the proposed method can deal with the evaluation level of attribute evaluation information and uncertainty of evaluation level trust.