diff --git a/tensorflow/python/eager/pywrap_tensor.cc b/tensorflow/python/eager/pywrap_tensor.cc
index 3e921793747..b4422bc5394 100644
--- a/tensorflow/python/eager/pywrap_tensor.cc
+++ b/tensorflow/python/eager/pywrap_tensor.cc
@@ -443,7 +443,7 @@ typedef struct EagerTensor {
   // Status objects on different functions that operate on EagerTensor and need
   // to use a TF_Status object. However note that accesses to `status` are not
   // thread-safe.
-  TF_Status* status;
+  TF_Status status;
 
   // The eager Context (from eager/context.py) used by this Tensor.
   // This is currently used only to make sure context outlives TensorHandles.
@@ -503,7 +503,7 @@ int EagerTensor_init(EagerTensor* self, PyObject* args, PyObject* kwds) {
   self->handle_data = Py_None;
   Py_INCREF(Py_None);
   self->tensor_shape = Py_None;
-  self->status = TF_NewStatus();
+  self->status.status = tensorflow::Status::OK();
   self->dict = nullptr;
   self->weakreflist = nullptr;
   self->context = nullptr;
@@ -543,7 +543,6 @@ void EagerTensor_dealloc(EagerTensor* self) {
   // Needs to happen before any actual destruction.
   PyObject_ClearWeakRefs((PyObject*)self);
 
-  TF_DeleteStatus(self->status);
   Py_DECREF(self->handle_data);
   Py_DECREF(self->tensor_shape);
   // If an attribute dictionary has been created, release it. Note that this
@@ -579,21 +578,21 @@ static PyObject* EagerTensor_datatype_enum(EagerTensor* self) {
 // Getter for `_shape_tuple`.
 static PyObject* EagerTensor_shape_tuple(EagerTensor* self) {
   auto handle = self->handle;
-  int n = TFE_TensorHandleNumDims(handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, nullptr)) {
+  int n = TFE_TensorHandleNumDims(handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, nullptr)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
   PyObject* shape = PyTuple_New(n);
   if (PyErr_Occurred()) return nullptr;
   for (int i = 0; i < n; ++i) {
     PyObject* dim =
-        PyLong_FromLongLong(TFE_TensorHandleDim(handle, i, self->status));
-    if (MaybeRaiseExceptionFromTFStatus(self->status, nullptr) ||
+        PyLong_FromLongLong(TFE_TensorHandleDim(handle, i, &self->status));
+    if (MaybeRaiseExceptionFromTFStatus(&self->status, nullptr) ||
         dim == nullptr || PyTuple_SetItem(shape, i, dim) != 0) {
       // Cleanup self->status before returning.
-      TF_SetStatus(self->status, TF_OK, "");
+      self->status.status = tensorflow::Status::OK();
       Py_DECREF(shape);
       if (dim != nullptr) Py_DECREF(dim);
       PyErr_SetString(PyExc_RuntimeError, "Error while creating shape");
@@ -605,10 +604,10 @@ static PyObject* EagerTensor_shape_tuple(EagerTensor* self) {
 
 // Getter for `_rank`.
 static PyObject* EagerTensor_rank(EagerTensor* self) {
-  int num_dims = TFE_TensorHandleNumDims(self->handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, nullptr)) {
+  int num_dims = TFE_TensorHandleNumDims(self->handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, nullptr)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
 #if PY_MAJOR_VERSION < 3
@@ -621,10 +620,10 @@ static PyObject* EagerTensor_rank(EagerTensor* self) {
 // Getter for `_num_elements`.
 static PyObject* EagerTensor_num_elements(EagerTensor* self) {
   auto handle = self->handle;
-  int n = TFE_TensorHandleNumElements(handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, nullptr)) {
+  int n = TFE_TensorHandleNumElements(handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, nullptr)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
   return PyLong_FromLongLong(n);
@@ -670,10 +669,11 @@ static PyObject* EagerTensor_copy_to_device(EagerTensor* self, PyObject* args,
   // Note that this is a shallow copy and will share the underlying buffer
   // if copying to the same device.
   TFE_TensorHandle* handle = TFE_TensorHandleCopyToDevice(
-      self->handle, GetContextHandle(self->context), device_name, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, PyExc_RuntimeError)) {
+      self->handle, GetContextHandle(self->context), device_name,
+      &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, PyExc_RuntimeError)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
 
@@ -686,11 +686,11 @@ static PyObject* EagerTensor_copy_to_device(EagerTensor* self, PyObject* args,
 // other.
 // Note that if `self` is not on CPU, we raise an Exception.
 static PyObject* EagerTensor_numpy_internal(EagerTensor* self) {
-  auto* py_array = TFE_TensorHandleToNumpy(self->handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, nullptr)) {
+  auto* py_array = TFE_TensorHandleToNumpy(self->handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, nullptr)) {
     Py_XDECREF(py_array);
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   } else {
     return PyArray_Return(reinterpret_cast<PyArrayObject*>(py_array));
@@ -699,10 +699,10 @@ static PyObject* EagerTensor_numpy_internal(EagerTensor* self) {
 
 // Getter `device`.
 static PyObject* EagerTensor_device(EagerTensor* self) {
-  const char* device = TFE_TensorHandleDeviceName(self->handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, PyExc_ValueError)) {
+  const char* device = TFE_TensorHandleDeviceName(self->handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, PyExc_ValueError)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
 #if PY_MAJOR_VERSION >= 3
@@ -715,10 +715,10 @@ static PyObject* EagerTensor_device(EagerTensor* self) {
 // Getter `backing_device`.
 static PyObject* EagerTensor_backing_device(EagerTensor* self) {
   const char* device =
-      TFE_TensorHandleBackingDeviceName(self->handle, self->status);
-  if (MaybeRaiseExceptionFromTFStatus(self->status, PyExc_ValueError)) {
+      TFE_TensorHandleBackingDeviceName(self->handle, &self->status);
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, PyExc_ValueError)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return nullptr;
   }
 #if PY_MAJOR_VERSION >= 3
@@ -779,10 +779,10 @@ static int EagerTensor_getbuffer(EagerTensor* self, Py_buffer* view,
   // DT_STRING so the following is only slightly slower than a NumPy-free
   // implementation.
   auto py_array = tensorflow::make_safe(
-      TFE_TensorHandleToNumpy(self->handle, self->status));
-  if (MaybeRaiseExceptionFromTFStatus(self->status, PyExc_BufferError)) {
+      TFE_TensorHandleToNumpy(self->handle, &self->status));
+  if (MaybeRaiseExceptionFromTFStatus(&self->status, PyExc_BufferError)) {
     // Cleanup self->status before returning.
-    TF_SetStatus(self->status, TF_OK, "");
+    self->status.status = tensorflow::Status::OK();
     return -1;
   }
   if (PyObject_GetBuffer(py_array.get(), view, flags) < 0) {
@@ -899,7 +899,7 @@ PyObject* EagerTensorFromHandle(TFE_TensorHandle* handle) {
     Py_INCREF(Py_None);
     t->tensor_shape = Py_None;
     t->handle = handle;
-    t->status = TF_NewStatus();
+    t->status.status = tensorflow::Status::OK();
     t->weakreflist = nullptr;
     PyObject* py_context = GetPyEagerContext();
     if (py_context == nullptr) {
@@ -927,17 +927,16 @@ tensorflow::DataType PyEagerTensor_Dtype(const PyObject* tensor) {
       reinterpret_cast<const EagerTensor*>(tensor)->handle));
 }
 
-tensorflow::int64 PyEagerTensor_NumElements(const PyObject* tensor) {
+tensorflow::int64 PyEagerTensor_NumElements(PyObject* tensor) {
   DCHECK(EagerTensor_CheckExact(tensor));
-  const EagerTensor* as_c_eager_tensor =
-      reinterpret_cast<const EagerTensor*>(tensor);
+  EagerTensor* as_c_eager_tensor = reinterpret_cast<EagerTensor*>(tensor);
   tensorflow::int64 result = TFE_TensorHandleNumElements(
-      as_c_eager_tensor->handle, as_c_eager_tensor->status);
+      as_c_eager_tensor->handle, &as_c_eager_tensor->status);
 
-  if (MaybeRaiseExceptionFromTFStatus(as_c_eager_tensor->status,
+  if (MaybeRaiseExceptionFromTFStatus(&as_c_eager_tensor->status,
                                       PyExc_ValueError)) {
     // Cleanup status before returning.
-    TF_SetStatus(as_c_eager_tensor->status, TF_OK, "");
+    as_c_eager_tensor->status.status = tensorflow::Status::OK();
     return -1;
   }
 
diff --git a/tensorflow/python/eager/pywrap_tensor.h b/tensorflow/python/eager/pywrap_tensor.h
index 0a462178e78..4c84b5ce6ea 100644
--- a/tensorflow/python/eager/pywrap_tensor.h
+++ b/tensorflow/python/eager/pywrap_tensor.h
@@ -23,7 +23,7 @@ limitations under the License.
 bool EagerTensor_CheckExact(const PyObject* o);
 tensorflow::int64 PyEagerTensor_ID(const PyObject* tensor);
 tensorflow::DataType PyEagerTensor_Dtype(const PyObject* tensor);
-tensorflow::int64 PyEagerTensor_NumElements(const PyObject* tensor);
+tensorflow::int64 PyEagerTensor_NumElements(PyObject* tensor);
 
 namespace tensorflow {