In this study, in order to consider the future efficiency of suppliers as well as their precedent efficiency in supplier selection process, first, the input, output and link of the suppliers were forecasted by adaptive neuro-fuzzy inference system (ANFIS). Then, the future and precedent efficiency of the suppliers were determined using the forecasted values and dynamic DEA. Then, the more efficient supplier was selected with regard to the efficiency concepts. Some further studies have been also taken for application of the recommended procedure, as using the confirmatory factor analysis (CFA), the criteria including price, delivery, quality, service, eco-costs and capacity of electronic trading were considered for selection of suppliers. The findings indicated that the recommended method has more accuracy and less error for forecasting the efficiency of suppliers. Also, two candidates have been selected eventually as the most efficient suppliers of the company.