diff --git a/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py b/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py
index 66a9c68a61c47d..4b9e862c71d862 100644
--- a/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py
+++ b/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py
@@ -739,6 +739,34 @@ def test_unicode_funcnames(self):
                 self.assertEqual(count, 1)
                 self.assert_samples_equal(samples, collector)
 
+    def test_surrogates_in_frame_names(self):
+        cases = [
+            ("\udce9", r"\udce9"),  # A surrogate-escaped filename byte.
+            ("\uddd0", r"\uddd0"),  # The surrogate from gh-160601.
+            ("\ud800", r"\ud800"),
+            ("\udfff", r"\udfff"),
+            ("\ud800\udfff", r"\ud800\udfff"),
+            ("caf\xe9\udce9\U0001f40d", "caf\xe9\\udce9\U0001f40d"),
+        ]
+        for name, escaped_name in cases:
+            with self.subTest(name=name):
+                frames = [
+                    make_frame(f"/path/{name}.py", 42, f"func_{name}"),
+                    make_frame("test.py", 1, "main"),
+                ]
+                samples = [[make_interpreter(0, [make_thread(1, frames)])]] * 3
+                collector, count = self.roundtrip(samples)
+                self.assertEqual(count, 3)
+
+                expected_frames = [
+                    make_frame(f"/path/{escaped_name}.py", 42,
+                               f"func_{escaped_name}"),
+                    frames[1],
+                ]
+                expected = [[make_interpreter(
+                    0, [make_thread(1, expected_frames)])]] * 3
+                self.assert_samples_equal(expected, collector)
+
     def test_special_char_filenames(self):
         """Filenames with special characters roundtrip exactly."""
         filenames = [
diff --git a/Misc/NEWS.d/next/Library/2026-10-24-18-00-00.gh-issue-160601.Utf8Bin.rst b/Misc/NEWS.d/next/Library/2026-10-24-18-00-00.gh-issue-160601.Utf8Bin.rst
new file mode 100644
index 00000000000000..b0b5264ea341c4
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2026-10-24-18-00-00.gh-issue-160601.Utf8Bin.rst
@@ -0,0 +1,3 @@
+Fix a :exc:`UnicodeEncodeError` in :mod:`profiling.sampling` binary mode
+when a sampled filename or function name contains surrogate characters.
+These characters are now written as backslash escapes in the binary file.
diff --git a/Modules/_remote_debugging/binary_io_writer.c b/Modules/_remote_debugging/binary_io_writer.c
index 1794017053da50..59dd953b6e88c4 100644
--- a/Modules/_remote_debugging/binary_io_writer.c
+++ b/Modules/_remote_debugging/binary_io_writer.c
@@ -393,12 +393,15 @@ writer_intern_string(BinaryWriter *writer, PyObject *string, uint32_t *index)
         }
     }
 
-    Py_ssize_t str_len;
-    const char *str_data = PyUnicode_AsUTF8AndSize(string, &str_len);
-    if (!str_data) {
+    /* Frame names can contain surrogates (e.g. from undecodable filenames).
+     * Escape them to keep the string table valid UTF-8. */
+    PyObject *encoded = PyUnicode_AsEncodedString(string, "utf-8", "backslashreplace");
+    if (encoded == NULL) {
         return -1;
     }
+    Py_ssize_t str_len = PyBytes_GET_SIZE(encoded);
     if ((uintmax_t)str_len > UINT32_MAX) {
+        Py_DECREF(encoded);
         writer->state = BINARY_WRITER_LIMIT_REACHED;
         PyErr_Format(PyExc_OverflowError,
             "string length %zd exceeds binary format maximum %u",
@@ -408,10 +411,12 @@ writer_intern_string(BinaryWriter *writer, PyObject *string, uint32_t *index)
 
     char *str_copy = PyMem_Malloc(str_len + 1);
     if (!str_copy) {
+        Py_DECREF(encoded);
         PyErr_NoMemory();
         return -1;
     }
-    memcpy(str_copy, str_data, str_len + 1);
+    memcpy(str_copy, PyBytes_AS_STRING(encoded), str_len + 1);
+    Py_DECREF(encoded);
 
     *index = (uint32_t)writer->string_count;