diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py
index 8e405ad72..9fb7ff044 100644
--- a/youtube_dl/extractor/extractors.py
+++ b/youtube_dl/extractor/extractors.py
@@ -938,6 +938,7 @@ from .udemy import (
 from .udn import UDNEmbedIE
 from .digiteka import DigitekaIE
 from .unistra import UnistraIE
+from .universcience import UniverscienceIE
 from .uol import UOLIE
 from .uplynk import (
     UplynkIE,
diff --git a/youtube_dl/extractor/universcience.py b/youtube_dl/extractor/universcience.py
new file mode 100644
index 000000000..70f97220e
--- /dev/null
+++ b/youtube_dl/extractor/universcience.py
@@ -0,0 +1,97 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    xpath_text,
+    xpath_element,
+    xpath_attr,
+    clean_html,
+)
+
+
+class UniverscienceIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?universcience\.tv/video-(.*)-(?P<id>[0-9]+)\.html'
+    _TEST = {
+        'url': 'http://www.universcience.tv/video-haro-sur-les-loups-o-5466.html',
+        'info_dict': {
+            'id': '5466',
+            'duration': 1990,
+            'ext': 'mp4',
+            'title': 'Haro sur les loups ?',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'description': 'startswith:Face au retour',
+            'creator': 'Sylvie Allonneau',
+            'subtitles': {
+                'fr': [{
+                    'url': 'http://universcience-webtv2-videos-pad.brainsonic.com/1/20121217-100607/attachedFiles/subtitles/2c4a2240149dbb984edc8afefea23a7c.srt',
+                }],
+            },
+        }
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        APIKey = self._html_search_regex(r'["\']APIKey["\'](.*)["\'](?P<APIKey>.*)["\']', webpage, 'APIKey', group='APIKey')
+        url_get_media = 'http://universcience-webtv2-services-pad.brainsonic.com/rest/getMedia?APIKey=' + APIKey + '&byMediaId=' + video_id
+        xml = self._download_xml(url_get_media, video_id)
+        path_media = xpath_element(xml, './medias/media', fatal=True)
+
+        title = xpath_text(path_media, './title')
+        creator = xpath_text(path_media, './author')
+        duration = int_or_none(xpath_text(path_media, './length'))
+        description = clean_html(xpath_text(path_media, './description'))
+        thumbnail = xpath_text(path_media, './thumbnail_url')
+
+        subtitles = {}
+        subtitle_urls = path_media.findall('./subtitles/subtitle')
+        for subtitle in subtitle_urls:
+            lang = subtitle.get('lang')
+            subtitles[lang] = [{
+                'url': subtitle.text,
+            }]
+
+        formats = []
+        path_media_source = './medias/media/media_sources/media_source'
+        for media_source in xml.findall(path_media_source):
+            format_url = xpath_text(media_source, 'source', fatal=True)
+            media_label = xpath_attr(media_source, './streaming_type', 'label')
+            media_width = self._search_regex(
+                r'.* (\d*) x \d*', media_label, 'width', default='None', fatal=False)
+            media_height = self._search_regex(
+                r'.* \d* x (\d*)', media_label, 'height', default='None', fatal=False)
+            media_label = self._search_regex(
+                r'(.*) (\d* x \d*)', media_label, 'media_label', default=media_label, fatal=False)
+
+            if (media_label == 'HLS') or (media_label == 'm3u8'):
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hds', fatal=False))
+            else:
+                format_info = {
+                    'width': int_or_none(media_width),
+                    'height': int_or_none(media_height),
+                    'tbr': int_or_none(xpath_attr(media_source, './streaming_type', 'bitrate')),
+                    # 'vcodec': xpath_attr -> bug sur regexp?
+                    'vcodec': media_source.find('streaming_type').get('html5_codec'),
+                    'url': format_url,
+                    'format_id': 'http-%s' % media_label,
+                }
+                formats.append(format_info)
+
+        podcast_url = xpath_text(path_media, './podcast_url')
+        formats.append({'format_id': 'podcast', 'vcodec': 'none', 'url': podcast_url})
+
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'duration': duration,
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'formats': formats,
+            'creator': creator,
+            'subtitles': subtitles,
+        }