diff --git a/pyproject.toml b/pyproject.toml
index 7ad5b26abc..c3647291a5 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -65,6 +65,7 @@ extras = [
     "cartopy>=0.21.0",
     "dask>=2020.12.0",
     "netCDF4>=1.7.1.post1",
+    "rioxarray>=0.13.0",
     "shapely>=1.6.4",
     "boto3>=1.26.45"
 ]
diff --git a/src/metpy/xarray.py b/src/metpy/xarray.py
index 5d5ae524d3..974a2152c7 100644
--- a/src/metpy/xarray.py
+++ b/src/metpy/xarray.py
@@ -746,6 +746,63 @@ def assign_y_x(self, force=False, tolerance=None):
         new_dataarray = self._data_array.assign_coords(**{y.name: y, x.name: x})
         return new_dataarray.metpy.assign_coordinates(None)
 
+    def to_geotiff(self, path, **kwargs):
+        """Write this DataArray to a GeoTIFF file, preserving the MetPy CRS.
+
+        This is a convenience bridge to ``rioxarray``: it attaches the DataArray's MetPy
+        coordinate reference system (as assigned by ``.parse_cf`` or ``.assign_crs``) and its
+        x/y spatial dimension coordinates so the field is written as a properly georeferenced
+        raster that GIS software (e.g. QGIS, ArcGIS, GDAL-based tooling) can read directly.
+
+        Parameters
+        ----------
+        path : str or path-like
+            Destination path for the output GeoTIFF.
+        kwargs
+            Additional keyword arguments passed through to ``rioxarray``'s ``to_raster``
+            (e.g. ``driver``, ``compress``, ``dtype``, ``tiled``).
+
+        Notes
+        -----
+        Requires the optional ``rioxarray`` dependency. A valid CRS coordinate must be present
+        (as assigned by ``.parse_cf`` or ``.assign_crs``), and the DataArray must have
+        identifiable x and y dimension coordinates. PyProj is used to communicate the CRS to
+        ``rioxarray``.
+
+        """
+        try:
+            import rioxarray  # noqa: F401
+        except ImportError:
+            raise ImportError(
+                'rioxarray is required to write GeoTIFF files. Install it with '
+                '`pip install rioxarray` or `conda install -c conda-forge rioxarray`.'
+            ) from None
+
+        # Identify the spatial dimension coordinates that rioxarray needs
+        try:
+            x, y = self.coordinates('x', 'y')
+        except AttributeError:
+            raise ValueError(
+                'Both x and y dimension coordinates are required to write a GeoTIFF. Verify '
+                'that the data has been parsed by MetPy with proper x and y dimension '
+                'coordinates.'
+            ) from None
+
+        # Grab the CRS before we drop the coordinate holding it
+        crs = self.pyproj_crs
+
+        # rioxarray/rasterio operate on plain arrays, so drop pint units first. Also drop the
+        # metpy_crs coordinate, which holds a CFProjection object that cannot be serialized as
+        # a raster tag--the CRS is instead communicated to rioxarray explicitly below.
+        data = self.dequantify()
+        if 'metpy_crs' in data.coords:
+            data = data.drop_vars('metpy_crs')
+
+        # Tell rioxarray which dimensions are spatial and what the CRS is, then write
+        data = data.rio.set_spatial_dims(x_dim=x.name, y_dim=y.name, inplace=False)
+        data = data.rio.write_crs(crs, inplace=False)
+        data.rio.to_raster(path, **kwargs)
+
 
 @xr.register_dataset_accessor('metpy')
 class MetPyDatasetAccessor:
diff --git a/tests/test_xarray.py b/tests/test_xarray.py
index e9fbb3cfdf..973ae789d4 100644
--- a/tests/test_xarray.py
+++ b/tests/test_xarray.py
@@ -74,6 +74,49 @@ def test_pyproj_projection(test_var):
     assert proj.coordinate_operation.method_name == 'Lambert Conic Conformal (1SP)'
 
 
+def test_to_geotiff(test_var, tmp_path):
+    """Test writing a parsed DataArray to a georeferenced GeoTIFF."""
+    rioxarray = pytest.importorskip('rioxarray')
+
+    # Reduce to a single 2D spatial field so we have a plain (y, x) raster
+    field = test_var
+    for dim in field.dims:
+        if dim not in ('y', 'x'):
+            field = field.isel({dim: 0})
+
+    path = tmp_path / 'test.tif'
+    field.metpy.to_geotiff(path)
+    assert path.exists()
+
+    # Read back and verify the georeferencing round-trips
+    result = rioxarray.open_rasterio(path)
+    assert result.rio.crs == pyproj.CRS(test_var.metpy.pyproj_crs)
+    assert result.rio.width == test_var.sizes['x']
+    assert result.rio.height == test_var.sizes['y']
+
+
+def test_to_geotiff_no_rioxarray(test_var, monkeypatch):
+    """Test that a helpful error is raised when rioxarray is not installed."""
+    import builtins
+    real_import = builtins.__import__
+
+    def fake_import(name, *args, **kwargs):
+        if name == 'rioxarray':
+            raise ImportError('No module named rioxarray')
+        return real_import(name, *args, **kwargs)
+
+    monkeypatch.setattr(builtins, '__import__', fake_import)
+    with pytest.raises(ImportError, match='rioxarray is required'):
+        test_var.metpy.to_geotiff('unused.tif')
+
+
+def test_to_geotiff_no_spatial_coords(test_ds_generic, tmp_path):
+    """Test that writing data without x/y dimension coordinates raises a helpful error."""
+    pytest.importorskip('rioxarray')
+    with pytest.raises(ValueError, match='x and y dimension coordinates are required'):
+        test_ds_generic['test'].metpy.to_geotiff(tmp_path / 'unused.tif')
+
+
 def test_no_projection(test_ds):
     """Test getting the crs attribute when not available produces a sensible error."""
     var = test_ds.lat