Similarity measurement is a critical component in any case-based reasoning (CBR) system. CBR is a superior technique for solving new problems based on previous experiences. Main assumption in CBR relies on the hypothesis that states similar problems should have similar solutions. This paper describes a comparative analysis on several commonly used similarity measures (Canberra, Clark, and Normalized Euclidean distance) in retrieving phase of the case-based reasoning approach to facilitate supplier selection. In addition, the proposed agent-based supplier selection framework was designed to use customer’s defined weights to evaluate the price, volume, quality grade, and delivery date of supply materials, and also provide them with alternative products which are closest to their first order if it was out of stock. Finally, based on the proposed framework, a numerical example of the used approach is illustrated.