SUMMARYSeveral methods have been proposed for synthesizing communication protocol specifications starting from given service specifications. Because of the inherently sequential nature of a finite state machine (FSM), some FSM‐based protocol synthesis methods assume that primitives in the service specifications cannot be executed simultaneously. However, other synthesis methods were introduced to handle controlled primitive concurrency by imposing restrictions on the applicable FSM topologies. This paper proposes two alternative FSM‐based protocol synthesis methods that eliminate the restrictions on concurrency imposed by earlier methods. The first method applies a sequential‐based synthesis method to derive a sequential protocol specification (P‐SPEC) from a service specification (S‐SPEC) and then applies several state‐expansion rules to remodel the resulting P‐SPEC to consider the concurrency behavior specified in the S‐SPEC. The second method remodels a concurrent S‐SPEC into a sequential‐like one by expanding its states and applies a sequential‐based synthesis method to derive the concurrent P‐SPEC. Thus, the paper's main contribution is proposing synthesis methods that allow the protocol designers to model their service specifications with concurrency behaviors, using FSM‐based models, and to derive, automatically, the corresponding protocol specifications for the concurrently executable protocol entities. The derived protocol specifications are guaranteed to be free of design errors; therefore, they do not require any further verification. The complexity of the two methods is discussed and their syntactic and semantic correctness are proven. As an example application, the synthesis method is used to derive the protocol specification of the H.323 call release standard used in Internet calls. Copyright © 2012 John Wiley & Sons, Ltd.