A major advantage of Service-Oriented Architecture (SOA) is the ability to enable rapid formation of large-scale distributed systems. A dominant factor for the success of service-oriented system is the responsiveness of services, i.e. the availability and timeliness of services when needed. As different levels of responsiveness are expected under various situations, a major limitation of most existing service-oriented systems is that they serve all requests in the order of their arrival. Hence, all clients receive the same level of responsiveness. In this paper, an approach to efficiently and transparently providing adaptive responsiveness is presented. This approach includes an extended SOA architecture with replication and situation-aware load balancing for enabling service providers to offer adaptive responsiveness to clients. Moreover, our approach provides adaptable system capacity at runtime which reduces the cost for service replications, such as energy consumption. Experimental results are also presented to demonstrate the advantages of our approach.