diff -r 13cefcbcc7da Lib/email/mime/text.py
--- a/Lib/email/mime/text.py	Thu Mar 22 16:24:33 2012 +0100
+++ b/Lib/email/mime/text.py	Thu Mar 22 15:36:03 2012 -0400
@@ -27,4 +27,13 @@
         """
         MIMENonMultipart.__init__(self, 'text', _subtype,
                                   **{'charset': _charset})
+
+        # If _charset was defualted, check to see see if there are non-ascii
+        # characters present. Default to utf-8 if there are.
+        if _charset =='us-ascii':
+            try:
+                _text.encode(_charset)
+            except UnicodeEncodeError:
+                _charset = 'utf-8'
+
         self.set_payload(_text, _charset)
diff -r 13cefcbcc7da Lib/test/test_email/test_email.py
--- a/Lib/test/test_email/test_email.py	Thu Mar 22 16:24:33 2012 +0100
+++ b/Lib/test/test_email/test_email.py	Thu Mar 22 15:36:03 2012 -0400
@@ -617,6 +617,19 @@
             abc
             """))
 
+    def test_unicode_body_defaults_to_utf8_encoding(self):
+        # Issue 14291
+        m = MIMEText('É testabc\n')
+        self.assertEqual(str(m),textwrap.dedent("""\
+            MIME-Version: 1.0
+            Content-Type: text/plain; charset="utf-8"
+            Content-Transfer-Encoding: base64
+
+            w4kgdGVzdGFiYwo=
+            """))
+
+
+
 # Test the email.encoders module
 class TestEncoders(unittest.TestCase):
 
@@ -642,7 +655,7 @@
         eq(msg['content-transfer-encoding'], '7bit')
         # Similar, but with 8bit data
         msg = MIMEText('hello \xf8 world')
-        eq(msg['content-transfer-encoding'], '8bit')
+        eq(msg['content-transfer-encoding'], 'base64')
         # And now with a different charset
         msg = MIMEText('hello \xf8 world', _charset='iso-8859-1')
         eq(msg['content-transfer-encoding'], 'quoted-printable')