In this paper, the vendor selection (VS) problem with fuzzy goals is considered. An interactive method is developed for solving multi-objective VS problems where fuzzy data are represented by using S-curve membership functions. The proposed method attempts simultaneously to minimize the total order costs, the number of rejected items and the number of late delivered items with reference to several constraints such as meeting buyers' demand, vendors' capacity, vendors' quota flexibility, vendors' allocated budget, etc. An interactive solution methodology is proposed to solve the multi-objective VS problem and to find a preferred compromise solution. Moreover, the performance of S-curve membership functions that represent uncertainty goals and constraints in VS problems with linear membership functions in an industrial case is compared. © 2010 Elsevier Ltd. All rights reserved.