diff --git a/w3af/core/controllers/misc/diff.py b/w3af/core/controllers/misc/diff.py
index 22b68e3f74..aa5b27494d 100644
--- a/w3af/core/controllers/misc/diff.py
+++ b/w3af/core/controllers/misc/diff.py
@@ -143,7 +143,7 @@ def chunked_diff(a, b):
     return ''.join(a_chunks), ''.join(b_chunks)
 
 
-def split_by_sep(seq):
+def split_by_sep(sequence):
     """
     This method will split the HTTP response body by various separators,
     such as new lines, tabs, <, double and single quotes.
@@ -178,7 +178,7 @@ def split_by_sep(seq):
     chunks without much meaning and reduce the performance improvement we
     have achieved.
 
-    :param seq: A string
+    :param sequence: A string which we will split
     :return: A list of strings (chunks) for the input string
     """
     #
@@ -195,5 +195,8 @@ def split_by_sep(seq):
     #
     # [0] https://github.com/andresriancho/w3af/blob/2ded693c959c91dc3e4daca276460d6c64ada479/w3af/core/controllers/misc/diff.py#L173
     #
-    translated_seq = string.translate(seq, TRANSLATION_TABLE)
+    try:
+        translated_seq = string.translate(sequence, TRANSLATION_TABLE)
+    except UnicodeDecodeError:
+        translated_seq = string.translate(sequence.encode('utf-8'), TRANSLATION_TABLE)
     return translated_seq.split('\0')
diff --git a/w3af/core/controllers/misc/tests/test_diff.py b/w3af/core/controllers/misc/tests/test_diff.py
index 76539fc09f..f3e2283518 100644
--- a/w3af/core/controllers/misc/tests/test_diff.py
+++ b/w3af/core/controllers/misc/tests/test_diff.py
@@ -135,6 +135,11 @@ def test_split_by_sep_2(self):
         result = split_by_sep('hello world<bye\nbye!')
         self.assertEqual(result, ['hello world', 'bye', 'bye!'])
 
+    def test_split_by_sep_utf8(self):
+        sequence = u'ąęż'
+        # this shouldn't rise UnicodeDecodeError
+        split_by_sep(sequence)
+
     def test_split_by_sep_perf(self):
         loops = 1000
         inputs = [unittest.__doc__,