diff --git a/cpp/benchmarks/replay/replay.cpp b/cpp/benchmarks/replay/replay.cpp
index df2b14846..65b680669 100644
--- a/cpp/benchmarks/replay/replay.cpp
+++ b/cpp/benchmarks/replay/replay.cpp
@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
  * SPDX-License-Identifier: Apache-2.0
  */
 
@@ -14,9 +14,8 @@
 #include <rmm/mr/owning_wrapper.hpp>
 #include <rmm/mr/pool_memory_resource.hpp>
 
+#include <cuda/iterator>
 #include <thrust/execution_policy.h>
-#include <thrust/iterator/constant_iterator.h>
-#include <thrust/iterator/discard_iterator.h>
 #include <thrust/reduce.h>
 
 #include <benchmark/benchmark.h>
@@ -277,8 +276,8 @@ std::vector<std::vector<rmm::detail::event>> parse_per_thread_events(std::string
     thrust::host,
     all_events.begin(),
     all_events.end(),
-    thrust::make_constant_iterator(1),
-    thrust::make_discard_iterator(),
+    cuda::make_constant_iterator(1),
+    cuda::make_discard_iterator(),
     std::back_inserter(events_per_thread),
     [](event const& lhs, event const& rhs) { return lhs.thread_id == rhs.thread_id; });
 
diff --git a/cpp/include/rmm/device_uvector.hpp b/cpp/include/rmm/device_uvector.hpp
index 95028a01c..b798cb04b 100644
--- a/cpp/include/rmm/device_uvector.hpp
+++ b/cpp/include/rmm/device_uvector.hpp
@@ -13,8 +13,8 @@
 #include <rmm/mr/per_device_resource.hpp>
 #include <rmm/resource_ref.hpp>
 
+#include <cuda/std/iterator>
 #include <cuda/std/span>
-#include <thrust/iterator/reverse_iterator.h>
 
 #include <cstddef>
 #include <type_traits>
@@ -80,9 +80,10 @@ class device_uvector {
   using iterator       = pointer;            ///< The type of the iterator returned by begin()
   using const_iterator = const_pointer;  ///< The type of the const iterator returned by cbegin()
   using reverse_iterator =
-    thrust::reverse_iterator<iterator>;  ///< The type of the iterator returned by rbegin()
+    cuda::std::reverse_iterator<iterator>;  ///< The type of the iterator returned by rbegin()
   using const_reverse_iterator =
-    thrust::reverse_iterator<const_iterator>;  ///< The type of the iterator returned by crbegin()
+    cuda::std::reverse_iterator<const_iterator>;  ///< The type of the iterator returned by
+                                                  ///< crbegin()
 
   RMM_EXEC_CHECK_DISABLE
   ~device_uvector() = default;
diff --git a/cpp/include/rmm/mr/cuda_async_managed_memory_resource.hpp b/cpp/include/rmm/mr/cuda_async_managed_memory_resource.hpp
index 5a3218cfd..3582108a2 100644
--- a/cpp/include/rmm/mr/cuda_async_managed_memory_resource.hpp
+++ b/cpp/include/rmm/mr/cuda_async_managed_memory_resource.hpp
@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2026, NVIDIA CORPORATION.
  * SPDX-License-Identifier: Apache-2.0
  */
 #pragma once
@@ -9,7 +9,6 @@
 #include <rmm/detail/error.hpp>
 #include <rmm/detail/export.hpp>
 #include <rmm/detail/runtime_capabilities.hpp>
-#include <rmm/detail/thrust_namespace.h>
 #include <rmm/mr/cuda_async_view_memory_resource.hpp>
 #include <rmm/mr/device_memory_resource.hpp>
 
diff --git a/cpp/include/rmm/mr/cuda_async_memory_resource.hpp b/cpp/include/rmm/mr/cuda_async_memory_resource.hpp
index da55ef5a6..8ddfa4d53 100644
--- a/cpp/include/rmm/mr/cuda_async_memory_resource.hpp
+++ b/cpp/include/rmm/mr/cuda_async_memory_resource.hpp
@@ -9,7 +9,6 @@
 #include <rmm/detail/error.hpp>
 #include <rmm/detail/export.hpp>
 #include <rmm/detail/runtime_capabilities.hpp>
-#include <rmm/detail/thrust_namespace.h>
 #include <rmm/mr/cuda_async_view_memory_resource.hpp>
 #include <rmm/mr/device_memory_resource.hpp>
 
diff --git a/cpp/include/rmm/mr/cuda_async_view_memory_resource.hpp b/cpp/include/rmm/mr/cuda_async_view_memory_resource.hpp
index 6cc3e06bf..08cb7e699 100644
--- a/cpp/include/rmm/mr/cuda_async_view_memory_resource.hpp
+++ b/cpp/include/rmm/mr/cuda_async_view_memory_resource.hpp
@@ -8,7 +8,6 @@
 #include <rmm/cuda_stream_view.hpp>
 #include <rmm/detail/error.hpp>
 #include <rmm/detail/export.hpp>
-#include <rmm/detail/thrust_namespace.h>
 #include <rmm/mr/device_memory_resource.hpp>
 
 #include <cuda_runtime_api.h>
diff --git a/cpp/include/rmm/mr/fixed_size_memory_resource.hpp b/cpp/include/rmm/mr/fixed_size_memory_resource.hpp
index 38f103f71..4bac00f8f 100644
--- a/cpp/include/rmm/mr/fixed_size_memory_resource.hpp
+++ b/cpp/include/rmm/mr/fixed_size_memory_resource.hpp
@@ -9,14 +9,12 @@
 #include <rmm/detail/error.hpp>
 #include <rmm/detail/export.hpp>
 #include <rmm/detail/logging_assert.hpp>
-#include <rmm/detail/thrust_namespace.h>
 #include <rmm/mr/detail/fixed_size_free_list.hpp>
 #include <rmm/mr/detail/stream_ordered_memory_resource.hpp>
 #include <rmm/resource_ref.hpp>
 
+#include <cuda/iterator>
 #include <cuda_runtime_api.h>
-#include <thrust/iterator/counting_iterator.h>
-#include <thrust/iterator/transform_iterator.h>
 
 #include <algorithm>
 #include <cstddef>
@@ -176,7 +174,7 @@ class fixed_size_memory_resource
       return block_type{static_cast<char*>(ptr) + index * block_size_};
     };
     auto first =
-      thrust::make_transform_iterator(thrust::make_counting_iterator(std::size_t{0}), block_gen);
+      cuda::make_transform_iterator(cuda::make_counting_iterator(std::size_t{0}), block_gen);
     return free_list(first, first + num_blocks);
   }
 
diff --git a/cpp/include/rmm/mr/pool_memory_resource.hpp b/cpp/include/rmm/mr/pool_memory_resource.hpp
index 82d94b946..ec0939bdf 100644
--- a/cpp/include/rmm/mr/pool_memory_resource.hpp
+++ b/cpp/include/rmm/mr/pool_memory_resource.hpp
@@ -10,7 +10,6 @@
 #include <rmm/detail/export.hpp>
 #include <rmm/detail/format.hpp>
 #include <rmm/detail/logging_assert.hpp>
-#include <rmm/detail/thrust_namespace.h>
 #include <rmm/logger.hpp>
 #include <rmm/mr/detail/coalescing_free_list.hpp>
 #include <rmm/mr/detail/stream_ordered_memory_resource.hpp>
@@ -20,8 +19,6 @@
 
 #include <cuda/std/type_traits>
 #include <cuda_runtime_api.h>
-#include <thrust/iterator/counting_iterator.h>
-#include <thrust/iterator/transform_iterator.h>
 
 #include <algorithm>
 #include <cstddef>
diff --git a/cpp/tests/logger_tests.cpp b/cpp/tests/logger_tests.cpp
index 9650af8c5..587a865e7 100644
--- a/cpp/tests/logger_tests.cpp
+++ b/cpp/tests/logger_tests.cpp
@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
  * SPDX-License-Identifier: Apache-2.0
  */
 
@@ -8,7 +8,7 @@
 #include <rmm/mr/cuda_memory_resource.hpp>
 #include <rmm/mr/logging_resource_adaptor.hpp>
 
-#include <thrust/iterator/zip_iterator.h>
+#include <cuda/iterator>
 
 #include <benchmarks/utilities/log_parser.hpp>
 #include <gtest/gtest.h>
@@ -96,10 +96,10 @@ void expect_log_events(std::string const& filename,
 {
   auto actual_events = rmm::detail::parse_csv(filename);
 
-  auto begin = thrust::make_zip_iterator(
-    cuda::std::make_tuple(expected_events.begin(), actual_events.begin()));
+  auto begin =
+    cuda::make_zip_iterator(cuda::std::make_tuple(expected_events.begin(), actual_events.begin()));
   auto end =
-    thrust::make_zip_iterator(cuda::std::make_tuple(expected_events.end(), actual_events.end()));
+    cuda::make_zip_iterator(cuda::std::make_tuple(expected_events.end(), actual_events.end()));
 
   std::for_each(begin, end, [](auto const& zipped) {
     auto [expected, actual] = zipped;