Context-adaptive services for overcoming various limitations of mobile environment have become important. They have the potential to maintain appropriate service levels according to the changing environment. However, most existing studies concentrate on adaptation modules in a single system: client, proxy, or server. Hence, these systems suffer from a concentrated workload when the number of users increases. This causes response time delays to user requests. This paper proposes a novel adaptation system that performs adaptation over distributed systems. The proposed system monitors the condition of each system using adaptation modules that are deployed on the client, proxy, and server, and selects the appropriate system on which to perform the adaptation. Therefore, the proposed system responds rapidly and stabilizes the system operation, even when overload is increased, such as by the increase in the number of user requests. We implemented a prototype of the system in order to evaluate it, and tested the distributed processing for multimedia content adaptation, simulating the overload on the server. We compared the proposed system with the existing adaptation approaches, in terms of the response time and system stability, and confirmed the effectiveness of the system through these experimental results.