diff --git a/doc/api.rst b/doc/api.rst
index 4af9bb01208..40913e2b721 100644
--- a/doc/api.rst
+++ b/doc/api.rst
@@ -292,6 +292,7 @@ DataArray contents
    DataArray.drop_vars
    DataArray.reset_coords
    DataArray.copy
+   DataArray.unique
 
 **ndarray methods**:
 :py:attr:`~DataArray.astype`
diff --git a/doc/whats-new.rst b/doc/whats-new.rst
index 77d6296acac..8be5c6b7d2f 100644
--- a/doc/whats-new.rst
+++ b/doc/whats-new.rst
@@ -61,6 +61,9 @@ New Features
   :py:class:`~core.groupby.DataArrayGroupBy`, inspired by pandas'
   :py:meth:`~pandas.core.groupby.GroupBy.get_group`.
   By `Deepak Cherian <https://github.com/dcherian>`_.
+- Implement :py:meth:`DataArray.unique` to return a flattened NumPy array
+  of unique values in the data array; convenient for iteration (:pull:`5091`).
+  By `Andrew Huang <https://github.com/ahuang11>`_.
 
 Breaking changes
 ~~~~~~~~~~~~~~~~
diff --git a/xarray/core/dataarray.py b/xarray/core/dataarray.py
index 8c5df5575ac..65f5c9689a5 100644
--- a/xarray/core/dataarray.py
+++ b/xarray/core/dataarray.py
@@ -4418,6 +4418,20 @@ def query(
         )
         return ds[self.name]
 
+    def unique(self):
+        """Return a raveled NumPy array of all the unique values in the
+        data array; can be convenient for iteration. Uniques are returned
+        in order of appearance. This does NOT sort.
+
+        Returns
+        -------
+        Returns
+        -------
+        ndarray
+            The unique values returned as a NumPy array..
+        """
+        return pd.unique(self.values.ravel())
+
     # this needs to be at the end, or mypy will confuse with `str`
     # https://mypy.readthedocs.io/en/latest/common_issues.html#dealing-with-conflicting-names
     str = utils.UncachedAccessor(StringAccessor)
diff --git a/xarray/tests/test_dataarray.py b/xarray/tests/test_dataarray.py
index 20e357543d6..8ca7a2cdc91 100644
--- a/xarray/tests/test_dataarray.py
+++ b/xarray/tests/test_dataarray.py
@@ -7259,3 +7259,15 @@ def test_deepcopy_obj_array():
     x0 = DataArray(np.array([object()]))
     x1 = deepcopy(x0)
     assert x0.values[0] is not x1.values[0]
+
+
+def test_unique_order():
+    expected = np.array([0, 2, 1, 3])
+    result = DataArray([0, 2, 1, 1, 3, 3, 3, 2]).unique()
+    assert (expected == result).all()
+
+
+def test_unique_2d():
+    expected = np.array([0, 2, 1, 3])
+    result = xr.DataArray([[0, 2, 1, 1], [3, 3, 3, 2]]).unique()
+    assert (expected == result).all()