diff --git a/mercurial/templatefilters.py b/mercurial/templatefilters.py
--- a/mercurial/templatefilters.py
+++ b/mercurial/templatefilters.py
@@ -197,7 +197,11 @@ def json(obj):
         return {None: 'null', False: 'false', True: 'true'}[obj]
     elif isinstance(obj, int) or isinstance(obj, float):
         return str(obj)
+    elif isinstance(obj, encoding.localstr):
+        u = encoding.fromlocal(obj).decode('utf-8')  # can round-trip
+        return '"%s"' % jsonescape(u)
     elif isinstance(obj, str):
+        # no encoding.fromlocal() because it may abort if obj can't be decoded
         u = unicode(obj, encoding.encoding, 'replace')
         return '"%s"' % jsonescape(u)
     elif isinstance(obj, unicode):
diff --git a/tests/test-command-template.t b/tests/test-command-template.t
--- a/tests/test-command-template.t
+++ b/tests/test-command-template.t
@@ -3479,3 +3479,26 @@ Test broken string escapes:
   $ hg log -T "\\xy" -R a
   hg: parse error: invalid \x escape
   [255]
+
+Set up repository for non-ascii encoding tests:
+
+  $ hg init nonascii
+  $ cd nonascii
+  $ python <<EOF
+  > open('utf-8', 'w').write('\xc3\xa9')
+  > EOF
+  $ HGENCODING=utf-8 hg branch -q `cat utf-8`
+  $ HGENCODING=utf-8 hg ci -qAm 'non-ascii branch' utf-8
+
+json filter should try round-trip conversion to utf-8:
+
+  $ HGENCODING=ascii hg log -T "{branch|json}\n" -r0
+  "\u00e9"
+
+json filter should not abort if it can't decode bytes:
+(not sure the current behavior is right; we might want to use utf-8b encoding?)
+
+  $ HGENCODING=ascii hg log -T "{'`cat utf-8`'|json}\n" -l1
+  "\ufffd\ufffd"
+
+  $ cd ..