Abstract The electronic trade between enterprises becomes more popular due to the rapid development of electronic commerce. Under the electronic business environment, in order to make the trade process smoother and faster, most enterprises adopt industry standards, such as EDI and RosettaNet, for developing electronic business platforms to transact business processes. These standards and platforms can help both trading sides to control and facilitate the business processes and validate the document format. However, these processes still need human efforts to make decisions, which might cause business losses due to time delay or human error during the process. In order to solve the problem, this study takes a step further to propose a negotiation framework for automating the B2B processes. This study chooses RosettaNet purchase order management processes as an example to implement automated negotiations on Biztalk platform. In order to evaluate the negotiation behavior of the opponent, two fuzzy membership functions, preference degree and opponent’s concession degree, are proposed to develop 25 levels of fuzzy rules to determine better offers to help to reach a settlement more responsively and dynamically. Experiment results show that our approach can help B2B transaction effectively.