Most existing trust negotiation languages can not simultaneously have the following important functions: Distributed trust proving, complicated access control definition and negotiation-related constraints. Based on RT (role-based trust-management) language, this paper proposes a distributed trust proving and negotiation orientated language RTP (role-based trust proving). It can support distributed trust proving, define complicated roles, protect the policy’s sensitive information and avoid unrelated credential fetching. Both the syntax and semantics of RTP are introduced. The paper also designs a distributed trust proving and negotiation algorithm based on RTP to demonstrate the efficiency of RTP. Experimental results show that the algorithm supports the functions aimed by RTP, and outperforms the traditional trust negotiation in terms of both time and number of credential transfers.