diff --git a/rdflib/plugins/parsers/notation3.py b/rdflib/plugins/parsers/notation3.py
index dedbfcc87..882d08664 100755
--- a/rdflib/plugins/parsers/notation3.py
+++ b/rdflib/plugins/parsers/notation3.py
@@ -384,6 +384,10 @@ def unicodeExpand(m: Match) -> str:
 langcode = re.compile(r"[a-zA-Z0-9]+(-[a-zA-Z0-9]+)*")
 
 
+class sfloat(str):  # noqa: N801
+    """don't normalize raw XSD.double string representation"""
+
+
 class SinkParser:
     def __init__(
         self,
@@ -1530,7 +1534,7 @@ def nodeOrLiteral(self, argstr: str, i: int, res: MutableSequence[Any]) -> int:
                 m = exponent_syntax.match(argstr, i)
                 if m:
                     j = m.end()
-                    res.append(float(argstr[i:j]))
+                    res.append(sfloat(argstr[i:j]))
                     return j
 
                 m = decimal_syntax.match(argstr, i)
@@ -1921,7 +1925,7 @@ def normalise(self, f: Formula | Graph | None, n: int) -> Literal: ...
     def normalise(self, f: Formula | Graph | None, n: Decimal) -> Literal: ...
 
     @overload
-    def normalise(self, f: Formula | Graph | None, n: float) -> Literal: ...
+    def normalise(self, f: Formula | Graph | None, n: sfloat) -> Literal: ...
 
     @overload
     def normalise(self, f: Formula | Graph | None, n: Node) -> Node: ...
@@ -1929,7 +1933,7 @@ def normalise(self, f: Formula | Graph | None, n: Node) -> Node: ...
     def normalise(
         self,
         f: Formula | Graph | None,
-        n: Union[tuple[int, str], bool, int, Decimal, float, Node, _AnyT],
+        n: Union[tuple[int, str], bool, int, Decimal, sfloat, Node, _AnyT],
     ) -> Union[URIRef, Literal, BNode, Node, _AnyT]:
         if isinstance(n, tuple):
             return URIRef(str(n[1]))
@@ -1949,7 +1953,7 @@ def normalise(
             s = Literal(value, datatype=DECIMAL_DATATYPE)
             return s
 
-        if isinstance(n, float):
+        if isinstance(n, sfloat):
             s = Literal(str(n), datatype=DOUBLE_DATATYPE)
             return s
 
@@ -1965,7 +1969,7 @@ def normalise(
         #    f.universals[n] = f.newBlankNode()
         #    return f.universals[n]
         # type error: Incompatible return value type (got "Union[int, _AnyT]", expected "Union[URIRef, Literal, BNode, _AnyT]")  [return-value]
-        return n  # type: ignore[return-value]
+        return n
 
     def intern(self, something: _AnyT) -> _AnyT:
         return something
diff --git a/test/test_n3.py b/test/test_n3.py
index d2c362a39..40f871868 100644
--- a/test/test_n3.py
+++ b/test/test_n3.py
@@ -251,6 +251,23 @@ def test_empty_prefix(self):
             g2
         ), "Document with declared empty prefix must match default #"
 
+    @pytest.mark.parametrize(
+        "do_normalize_literal, expected_result",
+        [(True, {"1.0", "10000000000.0"}), (False, {"1e10", "1e0"})],
+    )
+    def test_float_no_norm(self, do_normalize_literal, expected_result):
+        import rdflib
+
+        original_normalize_literal = rdflib.NORMALIZE_LITERALS
+        try:
+            rdflib.NORMALIZE_LITERALS = do_normalize_literal
+            g1 = Graph()
+            g1.parse(data=":a :b 1e10, 1e0 .", format="n3")
+            values = set(str(o) for o in g1.objects())
+            assert values == expected_result
+        finally:
+            rdflib.NORMALIZE_LITERALS = original_normalize_literal
+
 
 class TestRegularExpressions:
     def test_exponents(self):