diff --git a/satpy/readers/core/seviri.py b/satpy/readers/core/seviri.py
index 79af4d1052..e726437dd9 100644
--- a/satpy/readers/core/seviri.py
+++ b/satpy/readers/core/seviri.py
@@ -134,6 +134,15 @@
       scn['IR_108']['y'] = mi
       scn['IR_108'].sel(time=np.datetime64('2019-03-01T12:06:13.052000000'))
 
+  Computing this coordinate can add a performance penalty for some readers.
+  For ``seviri_l1b_native`` it can be disabled with:
+
+  .. code-block:: python
+
+      scene = satpy.Scene(filenames,
+                          reader='seviri_l1b_native',
+                          reader_kwargs={'include_scanline_acq_time': False})
+
 * HRIT and Native readers can add raw metadata from the file header, such
   as calibration coefficients, to dataset attributes. Use the reader keyword
   argument ``include_raw_metadata``. Here's an example for extracting
diff --git a/satpy/readers/seviri_l1b_native.py b/satpy/readers/seviri_l1b_native.py
index 1178db2f5b..ce1cfae268 100644
--- a/satpy/readers/seviri_l1b_native.py
+++ b/satpy/readers/seviri_l1b_native.py
@@ -173,11 +173,19 @@ class NativeMSGFileHandler(BaseFileHandler):
 
     See :mod:`satpy.readers.core.seviri`.
 
+    **Scanline acquisition time**
+
+    By default, datasets include the ``acq_time`` coordinate (mean acquisition
+    time per scanline). For native files this requires additional work and can
+    slow down loading. If not needed, disable it with
+    ``reader_kwargs={'include_scanline_acq_time': False}``.
+
     """
 
     def __init__(self, filename, filename_info, filetype_info,
                  calib_mode="nominal", fill_disk=False, ext_calib_coefs=None,
-                 include_raw_metadata=False, mda_max_array_size=100):
+                 include_raw_metadata=False, mda_max_array_size=100,
+                 include_scanline_acq_time=True):
         """Initialize the reader."""
         super(NativeMSGFileHandler, self).__init__(filename,
                                                    filename_info,
@@ -188,6 +196,7 @@ def __init__(self, filename, filename_info, filetype_info,
         self.fill_disk = fill_disk
         self.include_raw_metadata = include_raw_metadata
         self.mda_max_array_size = mda_max_array_size
+        self.include_scanline_acq_time = include_scanline_acq_time
 
         # Declare required variables.
         self.header = {}
@@ -583,7 +592,8 @@ def get_dataset(self, dataset_id, dataset_info):
             return None
 
         dataset = self.calibrate(xarr, dataset_id)
-        self._add_scanline_acq_time(dataset, dataset_id)
+        if self.include_scanline_acq_time:
+            self._add_scanline_acq_time(dataset, dataset_id)
         self._update_attrs(dataset, dataset_info)
 
         if self.fill_disk and not (dataset_id["name"] != "HRV" and self.mda["is_full_disk"]):
diff --git a/satpy/tests/reader_tests/test_seviri_l1b_native.py b/satpy/tests/reader_tests/test_seviri_l1b_native.py
index 5765024cc5..01db10fc1b 100644
--- a/satpy/tests/reader_tests/test_seviri_l1b_native.py
+++ b/satpy/tests/reader_tests/test_seviri_l1b_native.py
@@ -935,6 +935,7 @@ def file_handler(self):
             fh.ext_calib_coefs = {}
             fh.include_raw_metadata = False
             fh.mda_max_array_size = 100
+            fh.include_scanline_acq_time = True
         return fh
 
     @staticmethod
@@ -1007,6 +1008,24 @@ def test_get_dataset(self, file_handler):
         assert file_handler.end_time == dt.datetime(2006, 1, 1, 12, 30, 0)
         assert_attrs_equal(xarr.attrs, expected.attrs, tolerance=1e-4)
 
+    def test_get_dataset_without_scanline_acq_time(self, file_handler):
+        """Test optional skip of scanline acquisition-time coordinate."""
+        dataset_id = make_dataid(
+            name="VIS006",
+            resolution=3000,
+            calibration="counts"
+        )
+        dataset_info = {
+            "units": "1",
+            "wavelength": (1, 2, 3),
+            "standard_name": "counts"
+        }
+        file_handler.include_scanline_acq_time = False
+        with mock.patch.object(file_handler, "_add_scanline_acq_time") as mocked_add:
+            xarr = file_handler.get_dataset(dataset_id, dataset_info)
+        mocked_add.assert_not_called()
+        assert "acq_time" not in xarr.coords
+
     def test_time(self, file_handler):
         """Test start/end nominal/observation time handling."""
         assert dt.datetime(2006, 1, 1, 12, 15, 9, 304888) == file_handler.observation_start_time