diff --git a/include/grpc++/impl/codegen/completion_queue.h b/include/grpc++/impl/codegen/completion_queue.h
index 56864d6d536..d489a90c69c 100644
--- a/include/grpc++/impl/codegen/completion_queue.h
+++ b/include/grpc++/impl/codegen/completion_queue.h
@@ -222,9 +222,18 @@ class CompletionQueue : private GrpcLibraryCodegen {
 /// A specific type of completion queue used by the processing of notifications
 /// by servers. Instantiated by \a ServerBuilder.
 class ServerCompletionQueue : public CompletionQueue {
+ public:
+  bool IsFrequentlyPolled() { return is_frequently_polled_; }
+
  private:
+  bool is_frequently_polled_;
   friend class ServerBuilder;
-  ServerCompletionQueue() {}
+  /// \param is_frequently_polled Informs the GPRC library about whether the
+  /// server completion queue would be actively polled (by calling Next() or
+  /// AsyncNext()). By default all server completion queues are assumed to be
+  /// frequently polled.
+  ServerCompletionQueue(bool is_frequently_polled = true)
+      : is_frequently_polled_(is_frequently_polled) {}
 };
 
 }  // namespace grpc
diff --git a/include/grpc++/server_builder.h b/include/grpc++/server_builder.h
index 86c7fecef59..85af9aa57fb 100644
--- a/include/grpc++/server_builder.h
+++ b/include/grpc++/server_builder.h
@@ -102,7 +102,15 @@ class ServerBuilder {
   /// Add a completion queue for handling asynchronous services
   /// Caller is required to keep this completion queue live until
   /// the server is destroyed.
-  std::unique_ptr<ServerCompletionQueue> AddCompletionQueue();
+  ///
+  /// \param is_frequently_polled This is an optional parameter to inform GRPC
+  /// library about whether this completion queue would be frequently polled
+  /// (i.e by calling Next() or AsyncNext()). The default value is 'true' and is
+  /// the recommended setting. Setting this to 'false' (i.e not polling the
+  /// completion queue frequently) will have a significantly negative
+  /// performance impact and hence should not be used in production use cases.
+  std::unique_ptr<ServerCompletionQueue> AddCompletionQueue(
+      bool is_frequently_polled = true);
 
   /// Return a running server which is ready for processing calls.
   std::unique_ptr<Server> BuildAndStart();
diff --git a/include/grpc/grpc.h b/include/grpc/grpc.h
index 5c868aece37..059bd2ebc74 100644
--- a/include/grpc/grpc.h
+++ b/include/grpc/grpc.h
@@ -334,6 +334,15 @@ GRPCAPI void grpc_server_register_completion_queue(grpc_server *server,
                                                    grpc_completion_queue *cq,
                                                    void *reserved);
 
+/** Register a non-listening completion queue with the server. This API is
+    similar to grpc_server_register_completion_queue except that the server will
+    not use this completion_queue to listen to any incoming channels.
+
+    Registering a non-listening completion queue will have negative performance
+    impact and hence this API is not recommended for production use cases. */
+GRPCAPI void grpc_server_register_non_listening_completion_queue(
+    grpc_server *server, grpc_completion_queue *q, void *reserved);
+
 /** Add a HTTP2 over plaintext over tcp listener.
     Returns bound port number on success, 0 on failure.
     REQUIRES: server not started */
diff --git a/src/core/lib/surface/completion_queue.c b/src/core/lib/surface/completion_queue.c
index 5ec8808b508..f6f7ac880cc 100644
--- a/src/core/lib/surface/completion_queue.c
+++ b/src/core/lib/surface/completion_queue.c
@@ -70,6 +70,8 @@ struct grpc_completion_queue {
   int shutdown;
   int shutdown_called;
   int is_server_cq;
+  /** Can the server cq accept incoming channels */
+  int is_non_listening_server_cq;
   int num_pluckers;
   plucker pluckers[GRPC_MAX_COMPLETION_QUEUE_PLUCKERS];
   grpc_closure pollset_shutdown_done;
@@ -149,6 +151,7 @@ grpc_completion_queue *grpc_completion_queue_create(void *reserved) {
   cc->shutdown = 0;
   cc->shutdown_called = 0;
   cc->is_server_cq = 0;
+  cc->is_non_listening_server_cq = 0;
   cc->num_pluckers = 0;
 #ifndef NDEBUG
   cc->outstanding_tag_count = 0;
@@ -507,6 +510,14 @@ grpc_pollset *grpc_cq_pollset(grpc_completion_queue *cc) {
   return POLLSET_FROM_CQ(cc);
 }
 
+void grpc_cq_mark_non_listening_server_cq(grpc_completion_queue *cc) {
+  cc->is_non_listening_server_cq = 1;
+}
+
+bool grpc_cq_is_non_listening_server_cq(grpc_completion_queue *cc) {
+  return (cc->is_non_listening_server_cq == 1);
+}
+
 void grpc_cq_mark_server_cq(grpc_completion_queue *cc) { cc->is_server_cq = 1; }
 
 int grpc_cq_is_server_cq(grpc_completion_queue *cc) { return cc->is_server_cq; }
diff --git a/src/core/lib/surface/completion_queue.h b/src/core/lib/surface/completion_queue.h
index eef82cf0148..ee3e0448401 100644
--- a/src/core/lib/surface/completion_queue.h
+++ b/src/core/lib/surface/completion_queue.h
@@ -82,6 +82,8 @@ void grpc_cq_end_op(grpc_exec_ctx *exec_ctx, grpc_completion_queue *cc,
 
 grpc_pollset *grpc_cq_pollset(grpc_completion_queue *cc);
 
+void grpc_cq_mark_non_listening_server_cq(grpc_completion_queue *cc);
+bool grpc_cq_is_non_listening_server_cq(grpc_completion_queue *cc);
 void grpc_cq_mark_server_cq(grpc_completion_queue *cc);
 int grpc_cq_is_server_cq(grpc_completion_queue *cc);
 
diff --git a/src/core/lib/surface/server.c b/src/core/lib/surface/server.c
index cbfd2458741..c34ec04d2d3 100644
--- a/src/core/lib/surface/server.c
+++ b/src/core/lib/surface/server.c
@@ -895,23 +895,45 @@ const grpc_channel_filter grpc_server_top_filter = {
     "server",
 };
 
-void grpc_server_register_completion_queue(grpc_server *server,
-                                           grpc_completion_queue *cq,
-                                           void *reserved) {
+static void register_completion_queue(grpc_server *server,
+                                      grpc_completion_queue *cq,
+                                      bool is_non_listening, void *reserved) {
   size_t i, n;
-  GRPC_API_TRACE(
-      "grpc_server_register_completion_queue(server=%p, cq=%p, reserved=%p)", 3,
-      (server, cq, reserved));
   GPR_ASSERT(!reserved);
   for (i = 0; i < server->cq_count; i++) {
     if (server->cqs[i] == cq) return;
   }
-  GRPC_CQ_INTERNAL_REF(cq, "server");
+
   grpc_cq_mark_server_cq(cq);
-  n = server->cq_count++;
-  server->cqs = gpr_realloc(server->cqs,
-                            server->cq_count * sizeof(grpc_completion_queue *));
-  server->cqs[n] = cq;
+
+  /* Non-listening completion queues are not added to server->cqs */
+  if (is_non_listening) {
+    grpc_cq_mark_non_listening_server_cq(cq);
+  } else {
+    GRPC_CQ_INTERNAL_REF(cq, "server");
+    n = server->cq_count++;
+    server->cqs = gpr_realloc(
+        server->cqs, server->cq_count * sizeof(grpc_completion_queue *));
+    server->cqs[n] = cq;
+  }
+}
+
+void grpc_server_register_completion_queue(grpc_server *server,
+                                           grpc_completion_queue *cq,
+                                           void *reserved) {
+  GRPC_API_TRACE(
+      "grpc_server_register_completion_queue(server=%p, cq=%p, reserved=%p)", 3,
+      (server, cq, reserved));
+  register_completion_queue(server, cq, false, reserved);
+}
+
+void grpc_server_register_non_listening_completion_queue(
+    grpc_server *server, grpc_completion_queue *cq, void *reserved) {
+  GRPC_API_TRACE(
+      "grpc_server_register_non_listening_completion_queue(server=%p, cq=%p, "
+      "reserved=%p)",
+      3, (server, cq, reserved));
+  register_completion_queue(server, cq, true, reserved);
 }
 
 grpc_server *grpc_server_create(const grpc_channel_args *args, void *reserved) {
diff --git a/src/cpp/server/server_builder.cc b/src/cpp/server/server_builder.cc
index 68cc38258cf..5445d3e13bc 100644
--- a/src/cpp/server/server_builder.cc
+++ b/src/cpp/server/server_builder.cc
@@ -46,8 +46,9 @@ ServerBuilder::ServerBuilder()
   grpc_compression_options_init(&compression_options_);
 }
 
-std::unique_ptr<ServerCompletionQueue> ServerBuilder::AddCompletionQueue() {
-  ServerCompletionQueue* cq = new ServerCompletionQueue();
+std::unique_ptr<ServerCompletionQueue> ServerBuilder::AddCompletionQueue(
+    bool is_frequently_polled) {
+  ServerCompletionQueue* cq = new ServerCompletionQueue(is_frequently_polled);
   cqs_.push_back(cq);
   return std::unique_ptr<ServerCompletionQueue>(cq);
 }
@@ -105,8 +106,17 @@ std::unique_ptr<Server> ServerBuilder::BuildAndStart() {
   std::unique_ptr<Server> server(
       new Server(thread_pool.release(), true, max_message_size_, &args));
   for (auto cq = cqs_.begin(); cq != cqs_.end(); ++cq) {
-    grpc_server_register_completion_queue(server->server_, (*cq)->cq(),
-                                          nullptr);
+    // A completion queue that is not polled frequently (by calling Next() or
+    // AsyncNext()) is not safe to use for listening to incoming channels.
+    // Register all such completion queues as non-listening completion queues
+    // with the GRPC core library.
+    if ((*cq)->IsFrequentlyPolled()) {
+      grpc_server_register_completion_queue(server->server_, (*cq)->cq(),
+                                            nullptr);
+    } else {
+      grpc_server_register_non_listening_completion_queue(server->server_,
+                                                          (*cq)->cq(), nullptr);
+    }
   }
   for (auto service = services_.begin(); service != services_.end();
        service++) {
diff --git a/test/cpp/end2end/hybrid_end2end_test.cc b/test/cpp/end2end/hybrid_end2end_test.cc
index 02043a89d3a..0423448154d 100644
--- a/test/cpp/end2end/hybrid_end2end_test.cc
+++ b/test/cpp/end2end/hybrid_end2end_test.cc
@@ -216,7 +216,7 @@ class HybridEnd2endTest : public ::testing::Test {
     }
     // Create a separate cq for each potential handler.
     for (int i = 0; i < 5; i++) {
-      cqs_.push_back(builder.AddCompletionQueue());
+      cqs_.push_back(builder.AddCompletionQueue(false));
     }
     server_ = builder.BuildAndStart();
   }