We present CReMeS, a CORBA-compliant design and implementation of a new real-time communication service. It provides for efficient, predictable, and scalable communication between information producers and consumers. The CReMeS architecture is based on MidART’s Real-Time Channel-based Reflective Memory (RT-CRM) abstraction. This architecture supports the separation of QoS specification between producer and consumer of data and employs a userlevel scheduling scheme for communicating real-time tasks. These help us achieve end-to-end predictability and allows our service to scale. The CReMeS architecture provides a CORBA interface to applications and demands no changes to the ORB layer and the language mapping layer. Thus, it can run on non real-time Off-The-Shelf ORBs and enables applications on these ORBs to have scalable and end-to-end predictable asynchronous communication facility. In addition, an application designer can select whether to use an out-of-band channel or the ORB GIOP/IIOP for data communication. This permits a trade-off between performance, predictability and reliability. Experimental results demonstrate that our architecture can achieve better performance and predictability than a real-time implementation of the CORBA Event Service when the out-of-band channel is employed for data communication; it delivers better predictability with comparable performance when the ORB GIOP/IIOP is used.