diff --git a/app/models/searchgov_crawler.rb b/app/models/searchgov_crawler.rb
index edcaa35097..a3ec540c85 100644
--- a/app/models/searchgov_crawler.rb
+++ b/app/models/searchgov_crawler.rb
@@ -77,7 +77,7 @@ def skip_extensions_regex
 
   # avoid infinite loops caused by malformed urls
   def repeating_segments_regex
-    /(\/\w+)(?:.+?\1){2,}/
+    /(\/[[:alpha:]]+)(?=\/)(.*\1(?=\/)){2,}/
   end
 
   def supported_content_type(type)
diff --git a/spec/models/searchgov_crawler_spec.rb b/spec/models/searchgov_crawler_spec.rb
index 4fb550d19f..4500aa6cf8 100644
--- a/spec/models/searchgov_crawler_spec.rb
+++ b/spec/models/searchgov_crawler_spec.rb
@@ -316,4 +316,17 @@
       end
     end
   end
+
+  # Testing this private method directly for test speed
+  describe '.repeating_segments_regex' do
+    subject(:regex) { crawler.send(:repeating_segments_regex) }
+
+    it { is_expected.to match 'http://www.agency.gov/foo/foo/foo/' }
+    it { is_expected.to match 'http://www.agency.gov/foo/baz/foo/biz/foo/qux/' }
+    it { is_expected.not_to match 'http://www.agency.gov/fee/fie/foe/' }
+    it { is_expected.not_to match 'http://www.agency.gov/foo/foo/' }
+    it { is_expected.not_to match 'http://www.agency.gov/f/foo/foo' }
+    it { is_expected.not_to match 'http://www.agency.gov/foofoofoo/' }
+    it { is_expected.not_to match 'http://www.agency.gov/09/09/09/' }
+  end
 end