diff --git a/Lib/curses/textpad.py b/Lib/curses/textpad.py
index f6dfc990901d99..54d13cf72fbcbc 100644
--- a/Lib/curses/textpad.py
+++ b/Lib/curses/textpad.py
@@ -204,10 +204,8 @@ def gather(self):
             stop = self._end_of_line(y)
             if stop == 0 and self.stripspaces:
                 continue
-            for x in range(self.maxx+1):
-                if self.stripspaces and x > stop:
-                    break
-                result = result + str(self.win.in_wch(y, x))
+            count = stop+1 if self.stripspaces else self.maxx+1
+            result = result + str(self.win.in_wchstr(y, 0, count))
             if self.maxy > 0:
                 result = result + "\n"
         return result
diff --git a/Lib/test/test_curses.py b/Lib/test/test_curses.py
index ea2dcd76b585a9..32a39e60aaedfa 100644
--- a/Lib/test/test_curses.py
+++ b/Lib/test/test_curses.py
@@ -2660,6 +2660,22 @@ def test_textbox_combining(self):
                 box.do_command(ch)
             self.assertEqual(box.gather(), text + ' ')
 
+    @requires_wide_build
+    def test_textbox_double_width(self):
+        # A double-width (East Asian) character occupies two cells.  gather()
+        # reads a whole line at a time so that the second cell, which holds
+        # the same character, is not reported as another one.
+        text = '你好'
+        if self._encodable(text):
+            box, win = self._make_textbox(1, 12)
+            for ch in text:
+                box.do_command(ch)
+            self.assertEqual(box.gather(), text + ' ')
+            box, win = self._make_textbox(1, 12, stripspaces=0)
+            for ch in text:
+                box.do_command(ch)
+            self.assertEqual(box.gather(), text + ' ' * 8)
+
     def test_textbox_edit_wide(self):
         # edit() reads characters through get_wch().  Each character is pushed
         # with unget_wch(), which on a narrow build requires it to encode to a