diff --git a/doc/source/whatsnew/v1.4.0.rst b/doc/source/whatsnew/v1.4.0.rst
index e2b42a0ea14f5..5d487c4723a0d 100644
--- a/doc/source/whatsnew/v1.4.0.rst
+++ b/doc/source/whatsnew/v1.4.0.rst
@@ -377,6 +377,7 @@ I/O
 - Bug in :func:`read_fwf`, where difference in lengths of ``colspecs`` and ``names`` was not raising ``ValueError`` (:issue:`40830`)
 - Bug in :func:`Series.to_json` and :func:`DataFrame.to_json` where some attributes were skipped when serialising plain Python objects to JSON (:issue:`42768`, :issue:`33043`)
 - Column headers are dropped when constructing a :class:`DataFrame` from a sqlalchemy's ``Row`` object (:issue:`40682`)
+- Bug in unpickling a :class:`Index` with object dtype incorrectly inferring numeric dtypes (:issue:`43188`)
 -
 
 Period
diff --git a/pandas/core/indexes/base.py b/pandas/core/indexes/base.py
index cbb510d5669dc..38946935ce5bc 100644
--- a/pandas/core/indexes/base.py
+++ b/pandas/core/indexes/base.py
@@ -242,6 +242,10 @@ def _new_Index(cls, d):
             # GH#23752 "labels" kwarg has been replaced with "codes"
             d["codes"] = d.pop("labels")
 
+    elif "dtype" not in d and "data" in d:
+        # Prevent Index.__new__ from conducting inference;
+        #  "data" key not in RangeIndex
+        d["dtype"] = d["data"].dtype
     return cls.__new__(cls, **d)
 
 
diff --git a/pandas/tests/indexes/base_class/test_pickle.py b/pandas/tests/indexes/base_class/test_pickle.py
new file mode 100644
index 0000000000000..c670921decb78
--- /dev/null
+++ b/pandas/tests/indexes/base_class/test_pickle.py
@@ -0,0 +1,11 @@
+from pandas import Index
+import pandas._testing as tm
+
+
+def test_pickle_preserves_object_dtype():
+    # GH#43188, GH#43155 don't infer numeric dtype
+    index = Index([1, 2, 3], dtype=object)
+
+    result = tm.round_trip_pickle(index)
+    assert result.dtype == object
+    tm.assert_index_equal(index, result)