diff --git a/src/awkward/operations/ak_to_rdataframe.py b/src/awkward/operations/ak_to_rdataframe.py
index b64040df1f..617d5e2675 100644
--- a/src/awkward/operations/ak_to_rdataframe.py
+++ b/src/awkward/operations/ak_to_rdataframe.py
@@ -81,7 +81,7 @@ def _impl(
             none_policy="error",
             regulararray=True,
             use_from_iter=True,
-            primitive_policy="forbid",
+            primitive_policy="error",
             string_policy="as-characters",
         ).to_backend(cpu)
         if length is None:
diff --git a/tests/test_3375_add_rdf_column.py b/tests/test_3375_add_rdf_column.py
new file mode 100644
index 0000000000..faeddb09d0
--- /dev/null
+++ b/tests/test_3375_add_rdf_column.py
@@ -0,0 +1,91 @@
+# BSD 3-Clause License; see https://github.com/scikit-hep/awkward/blob/main/LICENSE
+
+from __future__ import annotations
+
+import numpy as np
+import pytest
+
+import awkward as ak
+import awkward._connect.cling
+import awkward._lookup
+
+ROOT = pytest.importorskip("ROOT")
+
+compiler = ROOT.gInterpreter.Declare
+
+
+# ---------------------------------------------------------------------
+def add_column(rdf: ROOT.RDataFrame, arr_val: np.ndarray, name: str):
+    v_col_org = rdf.GetColumnNames()
+    l_col_org = [name.c_str() for name in v_col_org]
+    l_col = []
+
+    for col in l_col_org:
+        l_col.append(col)
+
+    data = ak.from_rdataframe(rdf, columns=l_col)
+    d_data = {col: data[col] for col in l_col}
+
+    d_data[name] = arr_val
+
+    rdf = ak.to_rdataframe(d_data)
+
+    return rdf
+
+
+# ---------------------------------------------------------------------
+def test_add_column():
+    ak_array_x = ak.Array([1, 2, 3])
+    ak_array_y = ak.Array([4, 5, 6])
+
+    rdf = ak.to_rdataframe({"x": ak_array_x, "y": ak_array_y})
+
+    rdf = rdf.Define("z", "ROOT::RVec<int>({1, 2, 3})")
+    rdf = rdf.Define("w", "true")
+
+    arr_val = ak.Array([10, 20, 30])
+
+    rdf = add_column(rdf, arr_val, "values")
+
+    ROOT.ROOT.DisableImplicitMT()
+
+    rdf.Display().Print()
+
+
+# ---------------------------------------------------------------------
+def add_numpy_column(rdf: ROOT.RDataFrame, arr_val: np.ndarray, name: str):
+    v_col_org = rdf.GetColumnNames()
+    l_col_org = [name.c_str() for name in v_col_org]
+    l_col = []
+
+    for col in l_col_org:
+        l_col.append(col)
+
+    data = ak.from_rdataframe(rdf, columns=l_col)
+    d_data = {col: data[col] for col in l_col}
+
+    d_data[name] = arr_val
+
+    rdf = ak.to_rdataframe(d_data)
+
+    return rdf
+
+
+# ---------------------------------------------------------------------
+def test_add_numpy():
+    d_data = {
+        "x": np.array([1, 2, 3]),
+        "y": np.array([4, 5, 6]),
+    }
+
+    rdf = ROOT.RDF.FromNumpy(d_data)
+    rdf = rdf.Define("z", "ROOT::RVec<int>({1, 2, 3})")
+    rdf = rdf.Define("w", "true")
+
+    arr_val = np.array([10, 20, 30])
+
+    rdf = add_numpy_column(rdf, arr_val, "values")
+
+    ROOT.ROOT.DisableImplicitMT()
+
+    rdf.Display().Print()