diff --git a/scrapy/contrib/linkextractors/sgml.py b/scrapy/contrib/linkextractors/sgml.py
index 17877031d..03b2028cb 100644
--- a/scrapy/contrib/linkextractors/sgml.py
+++ b/scrapy/contrib/linkextractors/sgml.py
@@ -71,8 +71,8 @@ class BaseSgmlLinkExtractor(FixedSGMLParser):
         self.current_link = None
 
     def handle_data(self, data):
-        if self.current_link and not self.current_link.text:
-            self.current_link.text = data.strip()
+        if self.current_link:
+            self.current_link.text = self.current_link.text + data.strip()
 
     def matches(self, url):
         """This extractor matches with any url, since
diff --git a/scrapy/tests/test_contrib_linkextractors.py b/scrapy/tests/test_contrib_linkextractors.py
index a5e57bde0..0adc257ef 100644
--- a/scrapy/tests/test_contrib_linkextractors.py
+++ b/scrapy/tests/test_contrib_linkextractors.py
@@ -14,6 +14,7 @@ class LinkExtractorTestCase(unittest.TestCase):
         <p><a href="/about.html">About us</a></p>
         <img src="/logo.png" alt="Company logo (not a link)" />
         <p><a href="../othercat.html">Other category</a></p>
+        <p><a href="/">&gt;&gt;</a></p>
         <p><a href="/" /></p>
         </body></html>"""
         response = HtmlResponse("http://example.org/somepage/index.html", body=html)
@@ -23,6 +24,7 @@ class LinkExtractorTestCase(unittest.TestCase):
                          [Link(url='http://example.org/somepage/item/12.html', text='Item 12'), 
                           Link(url='http://example.org/about.html', text='About us'),
                           Link(url='http://example.org/othercat.html', text='Other category'), 
+                          Link(url='http://example.org/', text='>>'),
                           Link(url='http://example.org/', text='')])
 
     def test_base_url(self):