A system and method for application server with overload protection. A system, for example a server or a cluster, is designed to generate a message, and/or reject further work, whenever the server becomes overloaded. It does this in a processor-efficient manner so as not to place even more strain on the already burdened server. The rejection message or action is customizable. In accordance with an embodiment, the system determines where the request would appear in a priority queue, and rejects those further down the queue. It does this quickly and with minimal processing time by determining to which workload manager the request is going to be delivered.