diff --git a/Lib/test/test_str.py b/Lib/test/test_str.py
index 7bdd2881904548..b9ab13711db8c9 100644
--- a/Lib/test/test_str.py
+++ b/Lib/test/test_str.py
@@ -1736,8 +1736,6 @@ def __str__(self):
             'character buffers are decoded to unicode'
         )
 
-        self.assertRaises(TypeError, str, 42, 42, 42)
-
     def test_constructor_keyword_args(self):
         """Pass various keyword argument combinations to the constructor."""
         # The object argument can be passed as a keyword.
@@ -2652,22 +2650,45 @@ def test_check_encoding_errors(self):
         self.assertEqual(proc.rc, 10, proc)
 
     def test_str_invalid_call(self):
-        check = lambda *a, **kw: self.assertRaises(TypeError, str, *a, **kw)
-
         # too many args
-        check(1, "", "", 1)
+        with self.assertRaisesRegex(TypeError, r"str expected at most 3 arguments, got 4"):
+            str("too", "many", "argu", "ments")
+        with self.assertRaisesRegex(TypeError, r"str expected at most 3 arguments, got 4"):
+            str(1, "", "", 1)
 
         # no such kw arg
-        check(test=1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) got an unexpected keyword argument 'test'"):
+            str(test=1)
 
         # 'encoding' must be str
-        check(1, encoding=1)
-        check(1, 1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'encoding' must be str, not int"):
+            str(1, 1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'encoding' must be str, not int"):
+            str(1, encoding=1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'encoding' must be str, not bytes"):
+            str(b"x", b"ascii")
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'encoding' must be str, not bytes"):
+            str(b"x", encoding=b"ascii")
 
         # 'errors' must be str
-        check(1, errors=1)
-        check(1, "", errors=1)
-        check(1, 1, 1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'encoding' must be str, not int"):
+            str(1, 1, 1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'errors' must be str, not int"):
+            str(1, errors=1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'errors' must be str, not int"):
+            str(1, "", errors=1)
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'errors' must be str, not bytes"):
+            str(b"x", "ascii", b"strict")
+        with self.assertRaisesRegex(TypeError, r"str\(\) argument 'errors' must be str, not bytes"):
+            str(b"x", "ascii", errors=b"strict")
+
+        # both positional and kwarg
+        with self.assertRaisesRegex(TypeError, r"argument for str\(\) given by name \('encoding'\) and position \(2\)"):
+            str(b"x", "utf-8", encoding="ascii")
+        with self.assertRaisesRegex(TypeError, r"str\(\) takes at most 3 arguments \(4 given\)"):
+            str(b"x", "utf-8", "ignore", encoding="ascii")
+        with self.assertRaisesRegex(TypeError, r"str\(\) takes at most 3 arguments \(4 given\)"):
+            str(b"x", "utf-8", "strict", errors="ignore")
 
 
 class StringModuleTest(unittest.TestCase):
diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2024-08-10-12-44-03.gh-issue-122888.TUyu9r.rst b/Misc/NEWS.d/next/Core_and_Builtins/2024-08-10-12-44-03.gh-issue-122888.TUyu9r.rst
new file mode 100644
index 00000000000000..93171360df0dda
--- /dev/null
+++ b/Misc/NEWS.d/next/Core_and_Builtins/2024-08-10-12-44-03.gh-issue-122888.TUyu9r.rst
@@ -0,0 +1,2 @@
+Fix crash on certain calls to ``str()`` with positional arguments of the
+wrong type. Patch by Jelle Zijlstra.
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 12578812a762f6..da9c5857cc3bee 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -15121,7 +15121,16 @@ unicode_vectorcall(PyObject *type, PyObject *const *args,
         return PyObject_Str(object);
     }
     const char *encoding = arg_as_utf8(args[1], "encoding");
-    const char *errors = (nargs == 3) ? arg_as_utf8(args[2], "errors") : NULL;
+    if (encoding == NULL) {
+        return NULL;
+    }
+    const char *errors = NULL;
+    if (nargs == 3) {
+        errors = arg_as_utf8(args[2], "errors");
+        if (errors == NULL) {
+            return NULL;
+        }
+    }
     return PyUnicode_FromEncodedObject(object, encoding, errors);
 }