diff --git a/Lib/email/charset.py b/Lib/email/charset.py
--- a/Lib/email/charset.py
+++ b/Lib/email/charset.py
@@ -390,7 +390,9 @@
         if self.body_encoding is BASE64:
             if isinstance(string, str):
                 string = string.encode(self.output_charset)
-            return email.base64mime.body_encode(string)
+            cteinput = b'\r\n'.join(string.splitlines()) + (
+                            b'\r\n' if string[-1] in b'\r\n' else b'')
+            return email.base64mime.body_encode(cteinput)
         elif self.body_encoding is QP:
             return email.quoprimime.body_encode(string)
         else:
diff --git a/Lib/email/test/test_email.py b/Lib/email/test/test_email.py
--- a/Lib/email/test/test_email.py
+++ b/Lib/email/test/test_email.py
@@ -3304,7 +3304,7 @@
         Content-Type: text/plain; charset="utf-8"
         Content-Transfer-Encoding: base64
 
-        0JTQsCwg0L7QvdC4INC70LXRgtGP0YIuCg==
+        0JTQsCwg0L7QvdC4INC70LXRgtGP0YIuDQo=
         """)
 
     def test_generator_handles_8bit(self):
@@ -3920,6 +3920,40 @@
         c = Charset('fake')
         eq('hello world', c.body_encode('hello world'))
 
+    def test_base64_body_encode_normalizes_newlines(self):
+        # issue #975330.
+        c = Charset('utf-8')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\n world')),
+            b'hello\r\n world')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\n world\n')),
+            b'hello\r\n world\r\n')
+
+    def test_base64_body_encode_normalizes_os9_linesep(self):
+        # issue #975330.
+        c = Charset('utf-8')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\r world')),
+            b'hello\r\n world')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\r world\r')),
+            b'hello\r\n world\r\n')
+
+    def test_base64_body_encode_normalizes_mixed_linesep(self):
+        # issue #975330.
+        c = Charset('utf-8')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\rmy\r\nworld\n')),
+            b'hello\r\nmy\r\nworld\r\n')
+
+    def test_base64_body_encode_lack_of_final_linsep_preserved(self):
+        # issue #975330.
+        c = Charset('utf-8')
+        self.assertEqual(
+            email.base64mime.decode(c.body_encode('hello\rmy\r\nworld')),
+            b'hello\r\nmy\r\nworld')
+
     def test_unicode_charset_name(self):
         charset = Charset('us-ascii')
         self.assertEqual(str(charset), 'us-ascii')
diff --git a/Misc/NEWS b/Misc/NEWS
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -67,6 +67,10 @@
 Library
 -------
 
+- Issue #975330: when email uses base64mime to encode the text of a message
+  because of its charset, it now normalizes any line endings to the RFC
+  standard `\r\n` endings before doing the encoding.
+
 - Issue #14443: Tell rpmbuild to use the correct version of Python in
   bdist_rpm. Initial patch by Ross Lagerwall.