diff --git a/rollbar/lib/transforms/scruburl.py b/rollbar/lib/transforms/scruburl.py
index d14a505a..593fa9f4 100644
--- a/rollbar/lib/transforms/scruburl.py
+++ b/rollbar/lib/transforms/scruburl.py
@@ -42,7 +42,7 @@ def redact(self, url_string):
 
         try:
             url_parts = urlsplit(url_string)
-            qs_params = parse_qs(url_parts.query)
+            qs_params = parse_qs(url_parts.query, keep_blank_values=True)
         except:
             # This isn't a URL, return url_string which is a no-op
             # for this transform
diff --git a/rollbar/test/test_scruburl_transform.py b/rollbar/test/test_scruburl_transform.py
index 94cf4107..d5ab2e4d 100644
--- a/rollbar/test/test_scruburl_transform.py
+++ b/rollbar/test/test_scruburl_transform.py
@@ -49,7 +49,7 @@ def _compare_urls(self, url1, url2):
             url2 = '//%s' % url2
 
         parsed_urls = map(urlparse, (url1, url2))
-        qs_params = map(lambda x: parse_qs(x.query), parsed_urls)
+        qs_params = map(lambda x: parse_qs(x.query, keep_blank_values=True), parsed_urls)
         num_params = map(len, qs_params)
         param_names = map(lambda x: set(x.keys()), qs_params)
 
@@ -116,6 +116,11 @@ def test_scrub_missing_scheme_and_double_slash(self):
         expected = obj.replace('secr3t', '------').replace('secret', '------')
         self._assertScrubbed(['password'], obj, expected)
 
+    def test_keep_blank_url_params(self):
+        obj = 'http://foo.com/asdf?foo=bar&baz='
+        expected = obj
+        self._assertScrubbed(['password'], obj, expected, skip_id_check=True)
+
     def test_scrub_dict_val_isnt_string(self):
 
         # This link will *not* be scrubbed because the value isn't a string or bytes