diff --git a/Lib/csv.py b/Lib/csv.py
index 6cae34c705777d..fb4baf83ae3f5d 100644
--- a/Lib/csv.py
+++ b/Lib/csv.py
@@ -328,18 +328,17 @@ def _guess_quote_and_delimiter(self, data, delimiters):
             delim = ''
             skipinitialspace = 0
 
-        # if we see an extra quote between delimiters, we've got a
-        # double quoted format
+        # A doubled quote character inside a quoted field means
+        # a double quoted format.  Match whole fields, so that a match
+        # cannot slide across field boundaries.
         dq_regexp = re.compile(
-                               r"((%(delim)s)|^)\W*%(quote)s[^%(delim)s\n]*%(quote)s[^%(delim)s\n]*%(quote)s\W*((%(delim)s)|$)" % \
-                               {'delim':re.escape(delim), 'quote':quotechar}, re.MULTILINE)
-
-
-
-        if dq_regexp.search(data):
-            doublequote = True
-        else:
-            doublequote = False
+                r"(?:(?<=%(delim)s)|^) *+%(quote)s"           # ,"
+                r"((?:%(quote)s%(quote)s|[^%(quote)s]++)*+)"  # the body
+                r"%(quote)s(?:%(delim)s|$)"                   # ",
+                % {'delim': re.escape(delim), 'quote': quotechar},
+                re.MULTILINE)
+        dquotechar = quotechar * 2
+        doublequote = any(dquotechar in m[1] for m in dq_regexp.finditer(data))
 
         return (quotechar, doublequote, delim, skipinitialspace)
 
diff --git a/Lib/test/test_csv.py b/Lib/test/test_csv.py
index 2ab529b51c207d..8bb6bed36b4375 100644
--- a/Lib/test/test_csv.py
+++ b/Lib/test/test_csv.py
@@ -1564,6 +1564,25 @@ def test_zero_mode_tie_order_colon_first(self):
             sniffer.sniff(sample)
 
 
+    def test_sniff_regex_backtracking(self):
+        # gh-109638: this artificial sample used to take minutes.
+        sniffer = csv.Sniffer()
+        sample = '"",' * 100 + '"' * 100 + '0' + '"' * 100 + '0'
+        self.assertEqual(sniffer.sniff(sample).delimiter, ',')
+
+    def test_sniff_doublequote_across_fields(self):
+        # A quoted field which contains the delimiter, followed by
+        # an empty quoted field, is not a doubled quote.
+        sniffer = csv.Sniffer()
+        sample = '",","",","\n' * 4
+        dialect = sniffer.sniff(sample)
+        self.assertEqual(dialect.delimiter, ',')
+        self.assertEqual(dialect.quotechar, '"')
+        self.assertIs(dialect.doublequote, False)
+        self.assertEqual(next(csv.reader(StringIO(sample), dialect)),
+                         [',', '', ','])
+
+
 class NUL:
     def write(s, *args):
         pass
diff --git a/Misc/NEWS.d/next/Library/2026-07-29-11-25-00.gh-issue-109638.Vt2Rn9.rst b/Misc/NEWS.d/next/Library/2026-07-29-11-25-00.gh-issue-109638.Vt2Rn9.rst
new file mode 100644
index 00000000000000..f44163a20c8278
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2026-07-29-11-25-00.gh-issue-109638.Vt2Rn9.rst
@@ -0,0 +1,3 @@
+Fix exponential time in :meth:`csv.Sniffer.sniff` for a sample which contains
+many quote characters.  A doubled quote character is now also detected in
+a field which contains the delimiter or a line break.