diff --git a/doc/source/whatsnew/v1.2.0.rst b/doc/source/whatsnew/v1.2.0.rst
index 17cdb7538dad2..5479ac4df6afb 100644
--- a/doc/source/whatsnew/v1.2.0.rst
+++ b/doc/source/whatsnew/v1.2.0.rst
@@ -689,6 +689,7 @@ MultiIndex
 - Bug in :meth:`DataFrame.reset_index` with ``NaT`` values in index raises ``ValueError`` with message ``"cannot convert float NaN to integer"`` (:issue:`36541`)
 - Bug in :meth:`DataFrame.combine_first` when used with :class:`MultiIndex` containing string and ``NaN`` values raises ``TypeError`` (:issue:`36562`)
 - Bug in :meth:`MultiIndex.drop` dropped ``NaN`` values when non existing key was given as input (:issue:`18853`)
+- Bug in :meth:`MultiIndex.drop` dropping more values than expected when index has duplicates and is not sorted (:issue:`33494`)
 
 I/O
 ^^^
@@ -821,6 +822,7 @@ Other
 - Bug in :meth:`Index.intersection` with non-matching numeric dtypes casting to ``object`` dtype instead of minimal common dtype (:issue:`38122`)
 - Passing an array with 2 or more dimensions to the :class:`Series` constructor now raises the more specific ``ValueError`` rather than a bare ``Exception`` (:issue:`35744`)
 - Bug in ``dir`` where ``dir(obj)`` wouldn't show attributes defined on the instance for pandas objects (:issue:`37173`)
+- Bug in :meth:`Index.drop` raising ``InvalidIndexError`` when index has duplicates (:issue:`38051`)
 - Bug in :meth:`RangeIndex.difference` returning :class:`Int64Index` in some cases where it should return :class:`RangeIndex` (:issue:`38028`)
 
 .. ---------------------------------------------------------------------------
diff --git a/pandas/core/indexes/base.py b/pandas/core/indexes/base.py
index 40fcc824992b7..52ffb1567cb2d 100644
--- a/pandas/core/indexes/base.py
+++ b/pandas/core/indexes/base.py
@@ -5564,7 +5564,7 @@ def drop(self, labels, errors: str_t = "raise"):
         """
         arr_dtype = "object" if self.dtype == "object" else None
         labels = com.index_labels_to_array(labels, dtype=arr_dtype)
-        indexer = self.get_indexer(labels)
+        indexer = self.get_indexer_for(labels)
         mask = indexer == -1
         if mask.any():
             if errors != "ignore":
diff --git a/pandas/core/indexes/multi.py b/pandas/core/indexes/multi.py
index d575c67cb36aa..a28d33981bbbf 100644
--- a/pandas/core/indexes/multi.py
+++ b/pandas/core/indexes/multi.py
@@ -2165,7 +2165,8 @@ def drop(self, codes, level=None, errors="raise"):
                 if isinstance(loc, int):
                     inds.append(loc)
                 elif isinstance(loc, slice):
-                    inds.extend(range(loc.start, loc.stop))
+                    step = loc.step if loc.step is not None else 1
+                    inds.extend(range(loc.start, loc.stop, step))
                 elif com.is_bool_indexer(loc):
                     if self.lexsort_depth == 0:
                         warnings.warn(
diff --git a/pandas/tests/indexes/multi/test_drop.py b/pandas/tests/indexes/multi/test_drop.py
index c39954b22b0f2..f7b1bc4729428 100644
--- a/pandas/tests/indexes/multi/test_drop.py
+++ b/pandas/tests/indexes/multi/test_drop.py
@@ -1,3 +1,5 @@
+import warnings
+
 import numpy as np
 import pytest
 
@@ -149,6 +151,16 @@ def test_drop_with_nan_in_index(nulls_fixture):
         mi.drop(pd.Timestamp("2001"), level="date")
 
 
+def test_drop_with_non_monotonic_duplicates():
+    # GH#33494
+    mi = MultiIndex.from_tuples([(1, 2), (2, 3), (1, 2)])
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", PerformanceWarning)
+        result = mi.drop((1, 2))
+    expected = MultiIndex.from_tuples([(2, 3)])
+    tm.assert_index_equal(result, expected)
+
+
 def test_single_level_drop_partially_missing_elements():
     # GH 37820
 
diff --git a/pandas/tests/indexes/test_base.py b/pandas/tests/indexes/test_base.py
index ba49c51c9db8e..d5ca8a0f64fac 100644
--- a/pandas/tests/indexes/test_base.py
+++ b/pandas/tests/indexes/test_base.py
@@ -9,6 +9,7 @@
 import pytest
 
 from pandas._libs.tslib import Timestamp
+from pandas.compat import IS64
 from pandas.compat.numpy import np_datetime64_compat
 from pandas.util._test_decorators import async_mark
 
@@ -19,6 +20,7 @@
     DatetimeIndex,
     Float64Index,
     Int64Index,
+    IntervalIndex,
     PeriodIndex,
     RangeIndex,
     Series,
@@ -1505,6 +1507,17 @@ def test_drop_tuple(self, values, to_drop):
             with pytest.raises(KeyError, match=msg):
                 removed.drop(drop_me)
 
+    def test_drop_with_duplicates_in_index(self, index):
+        # GH38051
+        if len(index) == 0 or isinstance(index, MultiIndex):
+            return
+        if isinstance(index, IntervalIndex) and not IS64:
+            pytest.skip("Cannot test IntervalIndex with int64 dtype on 32 bit platform")
+        index = index.unique().repeat(2)
+        expected = index[2:]
+        result = index.drop(index[0])
+        tm.assert_index_equal(result, expected)
+
     @pytest.mark.parametrize(
         "attr",
         [