diff --git a/metric_learn/lmnn.py b/metric_learn/lmnn.py
index 600d55c0..7c0c2038 100644
--- a/metric_learn/lmnn.py
+++ b/metric_learn/lmnn.py
@@ -87,6 +87,12 @@ class LMNN(MahalanobisMixin, TransformerMixin):
       Tolerance of the optimization procedure. If the objective value varies
       less than `tol`, we consider the algorithm has converged and stop it.
 
+  use_pca : Not used
+
+      .. deprecated:: 0.5.0
+        `use_pca` was deprecated in version 0.5.0 and will
+        be removed in 0.6.0.
+
   verbose : bool, optional (default=False)
       Whether to print the progress of the optimization procedure.
 
@@ -151,7 +157,7 @@ class LMNN(MahalanobisMixin, TransformerMixin):
 
   def __init__(self, init=None, k=3, min_iter=50, max_iter=1000,
                learn_rate=1e-7, regularization=0.5, convergence_tol=0.001,
-               use_pca=True, verbose=False, preprocessor=None,
+               use_pca='deprecated', verbose=False, preprocessor=None,
                n_components=None, num_dims='deprecated', random_state=None):
     self.init = init
     self.k = k
@@ -173,6 +179,11 @@ def fit(self, X, y):
                     ' It has been deprecated in version 0.5.0 and will be'
                     ' removed in 0.6.0. Use "n_components" instead',
                     DeprecationWarning)
+    if self.use_pca != 'deprecated':
+      warnings.warn('"use_pca" parameter is not used.'
+                    ' It has been deprecated in version 0.5.0 and will be'
+                    ' removed in 0.6.0.',
+                    DeprecationWarning)
     k = self.k
     reg = self.regularization
     learn_rate = self.learn_rate
diff --git a/test/metric_learn_test.py b/test/metric_learn_test.py
index c49c9ef5..e05329ab 100644
--- a/test/metric_learn_test.py
+++ b/test/metric_learn_test.py
@@ -291,6 +291,18 @@ def test_changed_behaviour_warning(self):
       lmnn.fit(X, y)
     assert any(msg == str(wrn.message) for wrn in raised_warning)
 
+  def test_deprecation_use_pca(self):
+    # test that a DeprecationWarning is thrown about use_pca, if the
+    # default parameters are used.
+    # TODO: remove in v.0.6
+    X = np.array([[0, 0], [0, 1], [2, 0], [2, 1]])
+    y = np.array([1, 0, 1, 0])
+    lmnn = LMNN(k=2, use_pca=True)
+    msg = ('"use_pca" parameter is not used.'
+           ' It has been deprecated in version 0.5.0 and will be'
+           ' removed in 0.6.0.')
+    assert_warns_message(DeprecationWarning, msg, lmnn.fit, X, y)
+
 
 @pytest.mark.parametrize('X, y, loss', [(np.array([[0], [1], [2], [3]]),
                                          [1, 1, 0, 0], 3.0),
diff --git a/test/test_base_metric.py b/test/test_base_metric.py
index 313948ec..d3e7802c 100644
--- a/test/test_base_metric.py
+++ b/test/test_base_metric.py
@@ -26,7 +26,7 @@ def test_lmnn(self):
           "learn_rate=1e-07, "
           "max_iter=1000, min_iter=50, n_components=None, "
           "num_dims='deprecated', preprocessor=None, random_state=None, "
-          "regularization=0.5, use_pca=True, verbose=False)"))
+          "regularization=0.5, use_pca='deprecated', verbose=False)"))
 
   def test_nca(self):
     self.assertEqual(remove_spaces(str(metric_learn.NCA())),