diff --git a/picard/cluster.py b/picard/cluster.py
index 746f8ad15..78c4f40af 100644
--- a/picard/cluster.py
+++ b/picard/cluster.py
@@ -27,7 +27,6 @@ from picard.metadata import Metadata
 from picard.similarity import similarity2, similarity
 from picard.ui.item import Item
 from picard.util import format_time
-from picard.mbxml import artist_credit_from_node
 
 
 class Cluster(QtCore.QObject, Item):
diff --git a/picard/formats/asf.py b/picard/formats/asf.py
index 4ff5f4a43..e63a733a6 100644
--- a/picard/formats/asf.py
+++ b/picard/formats/asf.py
@@ -127,6 +127,7 @@ class ASFFile(File):
         'asin': 'ASIN',
         'djmixer': 'WM/DJMixer',
         'mixer': 'WM/Mixer',
+        'artists': 'WM/ARTISTS',
     }
     __RTRANS = dict([(b, a) for a, b in __TRANS.items()])
 
diff --git a/picard/formats/id3.py b/picard/formats/id3.py
index b63a85b69..27fc53380 100644
--- a/picard/formats/id3.py
+++ b/picard/formats/id3.py
@@ -156,6 +156,7 @@ class ID3File(File):
         'BARCODE': 'barcode',
         'ASIN': 'asin',
         'MusicMagic Fingerprint': 'musicip_fingerprint',
+        'Artists': 'artists',
     }
     __rtranslate_freetext = dict([(v, k) for k, v in __translate_freetext.iteritems()])
 
diff --git a/picard/formats/mp4.py b/picard/formats/mp4.py
index 410eb1b23..594a4cfd1 100644
--- a/picard/formats/mp4.py
+++ b/picard/formats/mp4.py
@@ -99,6 +99,7 @@ class MP4File(File):
         "----:com.apple.iTunes:MOOD": "mood",
         "----:com.apple.iTunes:SCRIPT": "script",
         "----:com.apple.iTunes:LANGUAGE": "language",
+        "----:com.apple.iTunes:ARTISTS": "artists",
     }
     __r_freeform_tags = dict([(v, k) for k, v in __freeform_tags.iteritems()])
 
diff --git a/picard/mbxml.py b/picard/mbxml.py
index 1838e4abe..9e79b13e7 100644
--- a/picard/mbxml.py
+++ b/picard/mbxml.py
@@ -135,26 +135,30 @@ def _translate_artist_node(node):
 def artist_credit_from_node(node):
     artist = ""
     artistsort = ""
+    artists = []
+    standardize_artists = config.setting["standardize_artists"]
     for credit in node.name_credit:
         a = credit.artist[0]
-        transl, translsort = _translate_artist_node(a)
-        if transl:
-            artist += transl
+        translated, translated_sort = _translate_artist_node(a)
+        if translated:
+            name = translated
+        elif 'name' in credit.children and not standardize_artists:
+            name = credit.name[0].text
         else:
-            if 'name' in credit.children and not config.setting["standardize_artists"]:
-                artist += credit.name[0].text
-            else:
-                artist += a.name[0].text
-        artistsort += translsort if translsort else a.sort_name[0].text
+            name = a.name[0].text
+        artist += name
+        artistsort += translated_sort if translated_sort else a.sort_name[0].text
+        artists.append(name)
         if 'joinphrase' in credit.attribs:
             artist += credit.joinphrase
             artistsort += credit.joinphrase
-    return (artist, artistsort)
+    return (artist, artistsort, artists)
 
 
 def artist_credit_to_metadata(node, m, release=False):
     ids = [n.artist[0].id for n in node.name_credit]
-    artist, artistsort = artist_credit_from_node(node)
+    artist, artistsort, artists = artist_credit_from_node(node)
+    m["artists"] = artists
     if release:
         m["musicbrainz_albumartistid"] = ids
         m["albumartist"] = artist
diff --git a/picard/util/tags.py b/picard/util/tags.py
index 9d099fa0d..a5705b1d0 100644
--- a/picard/util/tags.py
+++ b/picard/util/tags.py
@@ -86,6 +86,7 @@ TAG_NAMES = {
     'script': N_('Script'),
     '~length': N_('Length'),
     '~rating': N_('Rating'),
+    'artists': N_('Artists'),
 }
 
 
diff --git a/test/test_mbxml.py b/test/test_mbxml.py
index 495df5dc9..c0e7fd69c 100644
--- a/test/test_mbxml.py
+++ b/test/test_mbxml.py
@@ -2,7 +2,11 @@ import unittest
 import picard
 from picard import config
 from picard.metadata import Metadata
-from picard.mbxml import track_to_metadata, release_to_metadata
+from picard.mbxml import (
+    track_to_metadata,
+    release_to_metadata,
+    artist_credit_from_node
+)
 
 
 settings = {
@@ -135,3 +139,26 @@ class ReleaseTest(unittest.TestCase):
         self.assertEqual('B123456789', m['asin'])
         self.assertEqual('ABC', m['label'])
         self.assertEqual('ABC 123', m['catalognumber'])
+
+
+class ArtistTest(unittest.TestCase):
+
+    def test_1(self):
+        config.setting = settings
+        node = XmlNode(children={
+            'name_credit': [XmlNode(attribs={'joinphrase': ' & '}, children={
+                'artist': [XmlNode(attribs={'id': '456'}, children={
+                    'name': [XmlNode(text='Foo Bar')],
+                    'sort_name': [XmlNode(text='Bar, Foo')]
+                })]
+            }), XmlNode(children={
+                'artist': [XmlNode(attribs={'id': '789'}, children={
+                    'name': [XmlNode(text='Baz')],
+                    'sort_name': [XmlNode(text='Baz')]
+                })]
+            })]
+        })
+        artist, artist_sort, artists = artist_credit_from_node(node)
+        self.assertEqual(['Foo Bar', 'Baz'], artists)
+        self.assertEqual('Foo Bar & Baz', artist)
+        self.assertEqual('Bar, Foo & Baz', artist_sort)