diff --git a/python/cudf/cudf/core/column/string.py b/python/cudf/cudf/core/column/string.py
index 11dd7556812..de2df9b50d7 100644
--- a/python/cudf/cudf/core/column/string.py
+++ b/python/cudf/cudf/core/column/string.py
@@ -75,10 +75,6 @@
     is_space as cpp_isspace,
     is_upper as cpp_is_upper,
 )
-from cudf._lib.strings.convert.convert_integers import (
-    is_integer as cpp_is_integer,
-)
-from cudf._lib.strings.convert.convert_floats import is_float as cpp_is_float
 from cudf._lib.strings.combine import (
     concatenate as cpp_concatenate,
     join as cpp_join,
@@ -91,6 +87,10 @@
 from cudf._lib.strings.convert.convert_fixed_point import (
     to_decimal as cpp_to_decimal,
 )
+from cudf._lib.strings.convert.convert_floats import is_float as cpp_is_float
+from cudf._lib.strings.convert.convert_integers import (
+    is_integer as cpp_is_integer,
+)
 from cudf._lib.strings.convert.convert_urls import (
     url_decode as cpp_url_decode,
     url_encode as cpp_url_encode,
@@ -4760,10 +4760,7 @@ def base_size(self) -> int:
         if len(self.base_children) == 0:
             return 0
         else:
-            return int(
-                (self.base_children[0].size - 1)
-                / self.base_children[0].dtype.itemsize
-            )
+            return self.base_children[0].size - 1
 
     @property
     def data_array_view(self) -> cuda.devicearray.DeviceNDArray:
diff --git a/python/cudf/cudf/tests/test_serialize.py b/python/cudf/cudf/tests/test_serialize.py
index 656b66bf793..0e9c61b634d 100644
--- a/python/cudf/cudf/tests/test_serialize.py
+++ b/python/cudf/cudf/tests/test_serialize.py
@@ -296,3 +296,24 @@ def test_deserialize_cudf_0_16(datadir):
     actual = pickle.load(open(fname, "rb"))
 
     assert_eq(expected, actual)
+
+
+def test_serialize_sliced_string():
+    # https://github.com/rapidsai/cudf/issues/7735
+    data = ["hi", "hello", None]
+    pd_series = pd.Series(data, dtype=pd.StringDtype())
+    gd_series = cudf.Series(data, dtype="str")
+    sliced = gd_series[0:3]
+    serialized_gd_series = gd_series.serialize()
+    serialized_sliced = sliced.serialize()
+
+    # validate frames are equal or not
+    # because both should be identical
+    for i in range(3):
+        assert_eq(
+            serialized_gd_series[1][i].to_host_array(),
+            serialized_sliced[1][i].to_host_array(),
+        )
+
+    recreated = cudf.Series.deserialize(*sliced.serialize())
+    assert_eq(recreated.to_pandas(nullable=True), pd_series)
diff --git a/python/cudf/cudf/tests/test_string.py b/python/cudf/cudf/tests/test_string.py
index 98b8bfb870d..8b1ad696f04 100644
--- a/python/cudf/cudf/tests/test_string.py
+++ b/python/cudf/cudf/tests/test_string.py
@@ -2922,3 +2922,14 @@ def test_string_std():
     assert_exceptions_equal(
         lfunc=psr.std, rfunc=sr.std, compare_error_message=False
     )
+
+
+def test_string_slice_with_mask():
+    actual = cudf.Series(["hi", "hello", None])
+    expected = actual[0:3]
+
+    assert actual._column.base_size == 3
+    assert_eq(actual._column.base_size, expected._column.base_size)
+    assert_eq(actual._column.null_count, expected._column.null_count)
+
+    assert_eq(actual, expected)
diff --git a/python/dask_cudf/dask_cudf/tests/test_distributed.py b/python/dask_cudf/dask_cudf/tests/test_distributed.py
index cb3c696adc3..85354704902 100644
--- a/python/dask_cudf/dask_cudf/tests/test_distributed.py
+++ b/python/dask_cudf/dask_cudf/tests/test_distributed.py
@@ -6,11 +6,11 @@
 from dask.distributed import Client
 from distributed.utils_test import loop  # noqa: F401
 
-import dask_cudf
-
 import cudf
 from cudf.tests.utils import assert_eq
 
+import dask_cudf
+
 dask_cuda = pytest.importorskip("dask_cuda")
 
 
@@ -65,3 +65,13 @@ def test_ucx_seriesgroupby():
             dask_df_g = dask_df.groupby(["a"]).b.sum().compute()
 
             assert dask_df_g.name == "b"
+
+
+def test_str_series_roundtrip():
+    with dask_cuda.LocalCUDACluster(n_workers=1) as cluster:
+        with Client(cluster):
+            expected = cudf.Series(["hi", "hello", None])
+            dask_series = dask_cudf.from_cudf(expected, npartitions=2)
+
+            actual = dask_series.compute()
+            assert_eq(actual, expected)