Abstract: Aiming at the scheduling problem of flexible machining and assembly systems with reentrant subsystems, a negotiation framework was presented and a bargain-like bidirectional negotiation algorithm was developed in this paper. Multi-agents were adopted to construct the negotiation structure, which enable various agents have their own decision objectives. The objectives of job agents were to purchase appropriate services with low price, while the machine agents always try to sell their services with maximum profits. The global performance of scheduling solutions was improved since bidirectional negotiation provides a comprehensive decision rather than an independent local decision. Experimental results show that the standard deviation of total cost is 24%~207% lower than various combinations of dispatching rules. Therefore, the proposed algorithm has better performance and robustness.