diff --git a/include/nvexec/stream/bulk.cuh b/include/nvexec/stream/bulk.cuh
index a3a2a6ab8..9a885cf86 100644
--- a/include/nvexec/stream/bulk.cuh
+++ b/include/nvexec/stream/bulk.cuh
@@ -138,7 +138,7 @@ namespace nv::execution::_strm
         static_cast<Receiver&&>(rcvr),
         [&](_strm::opstate_base<Receiver>& stream_provider) -> receiver_t<Receiver>
         {
-          return receiver_t<Receiver>(self.shape_, static_cast<Fun&&>(self.fun_), stream_provider);
+          return receiver_t<Receiver>(self.shape_, static_cast<Self&&>(self).fun_, stream_provider);
         });
     }
     STDEXEC_EXPLICIT_THIS_END(connect)
diff --git a/test/nvexec/bulk.cpp b/test/nvexec/bulk.cpp
index 1573cb94a..35f5992f5 100644
--- a/test/nvexec/bulk.cpp
+++ b/test/nvexec/bulk.cpp
@@ -21,6 +21,14 @@ namespace
     (void) snd;
   }
 
+  TEST_CASE("nvexec bulk supports const lvalue senders", "[cuda][stream][adaptors][bulk]")
+  {
+    nvexec::stream_context stream_ctx{};
+    auto const snd = ex::schedule(stream_ctx.get_scheduler()) | ex::bulk(ex::par, 1, [](int) {});
+
+    REQUIRE(STDEXEC::sync_wait(snd).has_value());
+  }
+
   TEST_CASE("nvexec bulk executes on GPU", "[cuda][stream][adaptors][bulk]")
   {
     nvexec::stream_context stream_ctx{};