diff --git a/doc/user-guide/io.rst b/doc/user-guide/io.rst
index cf2b0b4b3ea..5de1e73f043 100644
--- a/doc/user-guide/io.rst
+++ b/doc/user-guide/io.rst
@@ -519,8 +519,9 @@ difference between values, with a reference time of the first time value.
 Coordinates
 ...........
 
-You can control the ``coordinates`` attribute written to disk by specifying ``DataArray.encoding["coordinates"]``.
-If not specified, xarray automatically sets ``DataArray.encoding["coordinates"]`` to a space-delimited list
+You can control the ``coordinates`` attribute written to disk by specifying the ``coordinates`` key in the
+output encoding keyword argument to `to_netcdf` or `to_zarr`.  If not specified, xarray automatically sets
+the ``coordinates`` attribute for each variable to a space-delimited list
 of names of coordinate variables that share dimensions with the ``DataArray`` being written.
 This allows perfect roundtripping of xarray datasets but may not be desirable.
 When an xarray ``Dataset`` contains non-dimensional coordinates that do not share dimensions with any of
diff --git a/doc/whats-new.rst b/doc/whats-new.rst
index 681fb1f070b..f133a28ee7e 100644
--- a/doc/whats-new.rst
+++ b/doc/whats-new.rst
@@ -36,7 +36,9 @@ Breaking changes
 
 Deprecations
 ~~~~~~~~~~~~
-
+- Setting encoding on Xarray objects directly via the ``encoding`` attribute is deprecated in favor of using the
+  ``encoding`` kwarg in :py:meth:`Dataset.to_netcdf` / :py:meth:`Dataset.to_zarr`. (:issue:`6323`, :pull:`7708`).
+  By `Joe Hamman <https://github.com/jhamman>`_.
 
 Bug fixes
 ~~~~~~~~~
diff --git a/xarray/backends/api.py b/xarray/backends/api.py
index e5adedbb576..3acdb7596dc 100644
--- a/xarray/backends/api.py
+++ b/xarray/backends/api.py
@@ -1254,7 +1254,7 @@ def dump_to_store(
     for k, enc in encoding.items():
         # no need to shallow copy the variable again; that already happened
         # in encode_dataset_coordinates
-        variables[k].encoding = enc
+        variables[k]._set_encoding_internal(enc)
         check_encoding.add(k)
 
     if encoder:
diff --git a/xarray/backends/store.py b/xarray/backends/store.py
index a507ee37470..06d829a58c1 100644
--- a/xarray/backends/store.py
+++ b/xarray/backends/store.py
@@ -58,7 +58,7 @@ def open_dataset(  # type: ignore[override]  # allow LSP violation, not supporti
         ds = Dataset(vars, attrs=attrs)
         ds = ds.set_coords(coord_names.intersection(vars))
         ds.set_close(filename_or_obj.close)
-        ds.encoding = encoding
+        ds._set_encoding_internal(encoding)
 
         return ds
 
diff --git a/xarray/backends/zarr.py b/xarray/backends/zarr.py
index 7d21c771e06..41d1a6a9e4b 100644
--- a/xarray/backends/zarr.py
+++ b/xarray/backends/zarr.py
@@ -594,7 +594,9 @@ def store(
             vars_with_encoding = {}
             for vn in existing_variable_names:
                 vars_with_encoding[vn] = variables[vn].copy(deep=False)
-                vars_with_encoding[vn].encoding = existing_vars[vn].encoding
+                vars_with_encoding[vn]._set_encoding_internal(
+                    existing_vars[vn].encoding
+                )
             vars_with_encoding, _ = self.encode(vars_with_encoding, {})
             variables_encoded.update(vars_with_encoding)
 
@@ -650,7 +652,7 @@ def set_variables(self, variables, check_encoding_set, writer, unlimited_dims=No
 
             fill_value = attrs.pop("_FillValue", None)
             if v.encoding == {"_FillValue": None} and fill_value is None:
-                v.encoding = {}
+                v._encoding = {}
 
             if name in self.zarr_group:
                 # existing variable
diff --git a/xarray/conventions.py b/xarray/conventions.py
index 780172879c6..74e011f51e8 100644
--- a/xarray/conventions.py
+++ b/xarray/conventions.py
@@ -694,7 +694,7 @@ def decode_cf(
     ds = Dataset(vars, attrs=attrs)
     ds = ds.set_coords(coord_names.union(extra_coords).intersection(vars))
     ds.set_close(close)
-    ds.encoding = encoding
+    ds._set_encoding_internal(encoding)
 
     return ds
 
diff --git a/xarray/core/alignment.py b/xarray/core/alignment.py
index 05e2ca7eb8b..1eaf5084f53 100644
--- a/xarray/core/alignment.py
+++ b/xarray/core/alignment.py
@@ -548,7 +548,7 @@ def _reindex_one(
             self.exclude_dims,
             self.exclude_vars,
         )
-        new_obj.encoding = obj.encoding
+        new_obj._set_encoding_internal(obj.encoding)
         return new_obj
 
     def reindex_all(self) -> None:
diff --git a/xarray/core/concat.py b/xarray/core/concat.py
index f092911948f..4e4efaf83b8 100644
--- a/xarray/core/concat.py
+++ b/xarray/core/concat.py
@@ -643,7 +643,7 @@ def get_indexes(name):
             f"Variables {absent_coord_names!r} are coordinates in some datasets but not others."
         )
     result = result.set_coords(coord_names)
-    result.encoding = result_encoding
+    result._set_encoding_internal(result_encoding)
 
     result = result.drop_vars(unlabeled_dims, errors="ignore")
 
diff --git a/xarray/core/dataarray.py b/xarray/core/dataarray.py
index ba55837ef7e..d5f32dfa586 100644
--- a/xarray/core/dataarray.py
+++ b/xarray/core/dataarray.py
@@ -870,13 +870,23 @@ def attrs(self, value: Mapping[Any, Any]) -> None:
     @property
     def encoding(self) -> dict[Any, Any]:
         """Dictionary of format-specific settings for how this array should be
-        serialized."""
+        serialized.
+
+        .. warning::
+            Setting encoding directly using the encoding property is deprecated.
+            Use the encoding kwarg in to_netcdf/to_zarr to set output encoding.
+        """
         return self.variable.encoding
 
     @encoding.setter
     def encoding(self, value: Mapping[Any, Any]) -> None:
+        # deprecated (variable setter will warn)
         self.variable.encoding = dict(value)
 
+    def _set_encoding_internal(self, value: Mapping[Any, Any]) -> None:
+        """temporary method to set encoding without issuing a FutureWarning"""
+        self.variable._set_encoding_internal(value)
+
     def reset_encoding(self: T_DataArray) -> T_DataArray:
         """Return a new DataArray without encoding on the array or any attached
         coords."""
diff --git a/xarray/core/dataset.py b/xarray/core/dataset.py
index d3a5ddd470e..d767ab96051 100644
--- a/xarray/core/dataset.py
+++ b/xarray/core/dataset.py
@@ -657,14 +657,32 @@ def attrs(self, value: Mapping[Any, Any]) -> None:
 
     @property
     def encoding(self) -> dict[Any, Any]:
-        """Dictionary of global encoding attributes on this dataset"""
+        """Dictionary of global encoding attributes on this dataset
+
+        .. warning::
+            Setting encoding directly using the encoding property is deprecated.
+            Use the encoding kwarg in to_netcdf/to_zarr to set output encoding.
+        """
         if self._encoding is None:
             self._encoding = {}
         return self._encoding
 
     @encoding.setter
     def encoding(self, value: Mapping[Any, Any]) -> None:
-        self._encoding = dict(value)
+        warnings.warn(
+            "Setting encoding directly using the encoding property is "
+            "deprecated. Use the encoding kwarg in to_netcdf/to_zarr to "
+            "set output encoding.",
+            category=FutureWarning,
+        )
+        self._set_encoding_internal(value)
+
+    def _set_encoding_internal(self, value: Mapping[Any, Any]) -> None:
+        """temporary method to set encoding without issuing a FutureWarning"""
+        try:
+            self._encoding = dict(value)
+        except ValueError:
+            raise ValueError("encoding must be castable to a dictionary")
 
     def reset_encoding(self: T_Dataset) -> T_Dataset:
         """Return a new Dataset without encoding on the dataset or any of its
@@ -2821,7 +2839,7 @@ def _reindex_callback(
             var = self._variables.get(name)
             if var is not None:
                 new_var.attrs = var.attrs
-                new_var.encoding = var.encoding
+                new_var._set_encoding_internal(var.encoding)
 
         # pass through indexes from excluded dimensions
         # no extra check needed for multi-coordinate indexes, potential conflicts
diff --git a/xarray/core/parallel.py b/xarray/core/parallel.py
index 2f8612c5a9b..d8d2edeee7a 100644
--- a/xarray/core/parallel.py
+++ b/xarray/core/parallel.py
@@ -553,7 +553,7 @@ def subset_dataset_to_block(
 
     for index in result._indexes:
         result[index].attrs = template[index].attrs
-        result[index].encoding = template[index].encoding
+        result[index]._set_encoding_internal(template[index].encoding)
 
     for name, gname_l in var_key_map.items():
         dims = template[name].dims
@@ -571,7 +571,7 @@ def subset_dataset_to_block(
             hlg, name=gname_l, chunks=var_chunks, dtype=template[name].dtype
         )
         result[name] = (dims, data, template[name].attrs)
-        result[name].encoding = template[name].encoding
+        result[name]._set_encoding_internal(template[name].encoding)
 
     result = result.set_coords(template._coord_names)
 
diff --git a/xarray/core/variable.py b/xarray/core/variable.py
index 9fe3c953aa6..cb4fe7b265c 100644
--- a/xarray/core/variable.py
+++ b/xarray/core/variable.py
@@ -365,7 +365,7 @@ def __init__(self, dims, data, attrs=None, encoding=None, fastpath=False):
         if attrs is not None:
             self.attrs = attrs
         if encoding is not None:
-            self.encoding = encoding
+            self._set_encoding_internal(encoding)
 
     @property
     def dtype(self):
@@ -969,13 +969,28 @@ def attrs(self, value: Mapping[Any, Any]) -> None:
 
     @property
     def encoding(self) -> dict[Any, Any]:
-        """Dictionary of encodings on this variable."""
+        """Dictionary of encodings on this variable.
+
+        .. warning::
+            Setting encoding directly using the encoding property is deprecated.
+            Use the encoding kwarg in to_netcdf/to_zarr to set output encoding.
+        """
         if self._encoding is None:
             self._encoding = {}
         return self._encoding
 
     @encoding.setter
-    def encoding(self, value):
+    def encoding(self, value: Mapping[Any, Any]) -> None:
+        warnings.warn(
+            "Setting encoding directly using the encoding property is "
+            "deprecated. Use the encoding kwarg in to_netcdf/to_zarr to "
+            "set output encoding.",
+            category=FutureWarning,
+        )
+        self._set_encoding_internal(value)
+
+    def _set_encoding_internal(self, value: Mapping[Any, Any]) -> None:
+        """temporary method to set encoding without issuing a FutureWarning"""
         try:
             self._encoding = dict(value)
         except ValueError:
diff --git a/xarray/tests/__init__.py b/xarray/tests/__init__.py
index 7e1b964ecba..919f5d0ee51 100644
--- a/xarray/tests/__init__.py
+++ b/xarray/tests/__init__.py
@@ -261,7 +261,7 @@ def create_test_data(seed: int | None = None, add_attrs: bool = True) -> Dataset
         "dim3",
         np.array([0, 1, 2, 0, 0, 1, 1, 2, 2, 3], dtype="int64"),
     )
-    obj.encoding = {"foo": "bar"}
+    obj._encoding = {"foo": "bar"}
     assert all(obj.data.flags.writeable for obj in obj.variables.values())
     return obj
 
diff --git a/xarray/tests/test_backends.py b/xarray/tests/test_backends.py
index 2b1e76774ca..e18b96eadab 100644
--- a/xarray/tests/test_backends.py
+++ b/xarray/tests/test_backends.py
@@ -512,6 +512,7 @@ def test_roundtrip_string_data(self) -> None:
         with self.roundtrip(expected) as actual:
             assert_identical(expected, actual)
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_roundtrip_string_encoded_characters(self) -> None:
         expected = Dataset({"x": ("t", ["ab", "cdef"])})
         expected["x"].encoding["dtype"] = "S1"
@@ -1504,6 +1505,7 @@ def test_encoding_kwarg_compression(self) -> None:
 
         assert ds.x.encoding == {}
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_keep_chunksizes_if_no_original_shape(self) -> None:
         ds = Dataset({"x": [1, 2, 3]})
         chunksizes = (2,)
@@ -1515,6 +1517,7 @@ def test_keep_chunksizes_if_no_original_shape(self) -> None:
                 ds["x"].encoding["chunksizes"], actual["x"].encoding["chunksizes"]
             )
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_encoding_chunksizes_unlimited(self) -> None:
         # regression test for GH1225
         ds = Dataset({"x": [1, 2, 3], "y": ("x", [2, 3, 4])})
@@ -1581,6 +1584,7 @@ def test_read_variable_len_strings(self) -> None:
                 with open_dataset(tmp_file, **cast(dict, kwargs)) as actual:
                     assert_identical(expected, actual)
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_encoding_unlimited_dims(self) -> None:
         ds = Dataset({"x": ("y", np.arange(10.0))})
         with self.roundtrip(ds, save_kwargs=dict(unlimited_dims=["y"])) as actual:
@@ -2855,6 +2859,7 @@ def test_cross_engine_read_write_netcdf3(self) -> None:
                                 for k in data.variables
                             ]
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_encoding_unlimited_dims(self) -> None:
         ds = Dataset({"x": ("y", np.arange(10.0))})
         with self.roundtrip(ds, save_kwargs=dict(unlimited_dims=["y"])) as actual:
@@ -2938,6 +2943,7 @@ def test_read_byte_attrs_as_unicode(self) -> None:
                 expected = Dataset(attrs={"foo": "bar"})
                 assert_identical(expected, actual)
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_encoding_unlimited_dims(self) -> None:
         ds = Dataset({"x": ("y", np.arange(10.0))})
         with self.roundtrip(ds, save_kwargs=dict(unlimited_dims=["y"])) as actual:
diff --git a/xarray/tests/test_coding.py b/xarray/tests/test_coding.py
index f7579c4b488..16a37dde678 100644
--- a/xarray/tests/test_coding.py
+++ b/xarray/tests/test_coding.py
@@ -56,6 +56,7 @@ def test_CFMaskCoder_encode_missing_fill_values_conflict(data, encoding) -> None
         assert_identical(roundtripped, original)
 
 
+@pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
 def test_CFMaskCoder_missing_value() -> None:
     expected = xr.DataArray(
         np.array([[26915, 27755, -9999, 27705], [25595, -9999, 28315, -9999]]),
diff --git a/xarray/tests/test_coding_times.py b/xarray/tests/test_coding_times.py
index 580de878fe6..9338da914b1 100644
--- a/xarray/tests/test_coding_times.py
+++ b/xarray/tests/test_coding_times.py
@@ -712,6 +712,7 @@ def test_decode_cf_time_bounds() -> None:
 
 
 @requires_cftime
+@pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
 def test_encode_time_bounds() -> None:
     time = pd.date_range("2000-01-16", periods=1)
     time_bounds = pd.date_range("2000-01-01", periods=2, freq="MS")
diff --git a/xarray/tests/test_concat.py b/xarray/tests/test_concat.py
index f60308f8863..84d05ba46e6 100644
--- a/xarray/tests/test_concat.py
+++ b/xarray/tests/test_concat.py
@@ -1015,6 +1015,7 @@ def test_concat(self) -> None:
         with pytest.raises(ValueError, match=r"not a valid argument"):
             concat([foo, bar], dim="w", data_vars="minimal")
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_concat_encoding(self) -> None:
         # Regression test for GH1297
         ds = Dataset(
diff --git a/xarray/tests/test_conventions.py b/xarray/tests/test_conventions.py
index 9485b506b89..17db1b17b20 100644
--- a/xarray/tests/test_conventions.py
+++ b/xarray/tests/test_conventions.py
@@ -154,6 +154,7 @@ def test_var_with_coord_attr(self) -> None:
         # Should not have any global coordinates.
         assert "coordinates" not in attrs
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_do_not_overwrite_user_coordinates(self) -> None:
         orig = Dataset(
             coords={"x": [0, 1, 2], "y": ("x", [5, 6, 7]), "z": ("x", [8, 9, 10])},
diff --git a/xarray/tests/test_dataarray.py b/xarray/tests/test_dataarray.py
index 4ffa95e16e6..aae467acb31 100644
--- a/xarray/tests/test_dataarray.py
+++ b/xarray/tests/test_dataarray.py
@@ -269,6 +269,7 @@ def test_sizes(self) -> None:
         with pytest.raises(TypeError):
             array.sizes["foo"] = 5  # type: ignore
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_encoding(self) -> None:
         expected = {"foo": "bar"}
         self.dv.encoding["foo"] = "bar"
@@ -278,6 +279,7 @@ def test_encoding(self) -> None:
         self.dv.encoding = expected2
         assert expected2 is not self.dv.encoding
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_reset_encoding(self) -> None:
         array = self.mda
         encoding = {"scale_factor": 10}
@@ -3324,6 +3326,7 @@ def test_series_categorical_index(self) -> None:
         arr = DataArray(s)
         assert "'a'" in repr(arr)  # should not error
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     @pytest.mark.parametrize("encoding", [True, False])
     def test_to_and_from_dict(self, encoding) -> None:
         array = DataArray(
@@ -6735,3 +6738,12 @@ def test_error_on_ellipsis_without_list(self) -> None:
         da = DataArray([[1, 2], [1, 2]], dims=("x", "y"))
         with pytest.raises(ValueError):
             da.stack(flat=...)  # type: ignore
+
+
+def test_setting_encoding_property_warns_deprecated():
+    da = DataArray([0, 1, 2])
+    with pytest.warns(
+        FutureWarning,
+        match=r"Setting encoding directly using the encoding property is deprecated.*",
+    ):
+        da.encoding = {"dtype": "f4"}
diff --git a/xarray/tests/test_dataset.py b/xarray/tests/test_dataset.py
index 422166234aa..5b9a07a2635 100644
--- a/xarray/tests/test_dataset.py
+++ b/xarray/tests/test_dataset.py
@@ -2873,6 +2873,7 @@ def test_copy_with_data_errors(self) -> None:
         with pytest.raises(ValueError, match=r"contain all variables in original"):
             orig.copy(data={"var1": new_var1})
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_reset_encoding(self) -> None:
         orig = create_test_data()
         vencoding = {"scale_factor": 10}
@@ -3037,6 +3038,7 @@ def test_rename_multiindex(self) -> None:
             with pytest.warns(UserWarning, match="does not create an index anymore"):
                 original.rename({"a": "b"})
 
+    @pytest.mark.filterwarnings("ignore:Setting encoding directly.*:FutureWarning")
     def test_rename_perserve_attrs_encoding(self) -> None:
         # test propagate attrs/encoding to new variable(s) created from Index object
         original = Dataset(coords={"x": ("x", [0, 1, 2])})
@@ -7024,3 +7026,18 @@ def test_transpose_error() -> None:
         ),
     ):
         ds.transpose(["y", "x"])  # type: ignore
+
+
+def test_setting_encoding_property_warns_deprecated():
+    ds = xr.Dataset({"foo": (("x", "y"), [[21]]), "bar": (("x", "y"), [[12]])})
+    with pytest.warns(
+        FutureWarning,
+        match=r"Setting encoding directly using the encoding property is deprecated.*",
+    ):
+        ds.encoding = {"unlimited_dims": {"x"}}
+
+    with pytest.warns(
+        FutureWarning,
+        match=r"Setting encoding directly using the encoding property is deprecated.*",
+    ):
+        ds["foo"].encoding = {"dtype": "f4"}
diff --git a/xarray/tests/test_variable.py b/xarray/tests/test_variable.py
index 2f571eb9a83..f3103ae4186 100644
--- a/xarray/tests/test_variable.py
+++ b/xarray/tests/test_variable.py
@@ -2963,3 +2963,12 @@ def test_pandas_two_only_timedelta_conversion_warning() -> None:
         var = Variable(["time"], data)
 
     assert var.dtype == np.dtype("timedelta64[ns]")
+
+
+def test_setting_encoding_property_warns_deprecated():
+    v = Variable("x", [1, 2, 3])
+    with pytest.warns(
+        FutureWarning,
+        match=r"Setting encoding directly using the encoding property is deprecated.*",
+    ):
+        v.encoding = {"dtype": "f4"}