diff --git a/TESTING.md b/TESTING.md
index d7616c9..c34daee 100644
--- a/TESTING.md
+++ b/TESTING.md
@@ -509,9 +509,11 @@ and skip the parse.
 Filters (`mode=workspace`; anything invalid is 400 "Invalid gallery filters"):
 
 - `q` (at most 500 characters). An entry passes when the case-folded relative path
-  contains it, or when its linked positive's text, Library title, a tag name, or
-  its pair title contains it (SQLite `LIKE` with escaping, ASCII case folding
-  only). Search does not apply the card title gate, and does not search pair tags.
+  contains it, or when its linked positive's text, Library title or a tag name
+  contains it (SQLite `LIKE` with escaping, ASCII case folding only). A pair's own
+  title names its negative variant, so it is not searched (#249 D12): a variant
+  name alone matches nothing. Search does not apply the card title gate, and does
+  not search pair tags.
 - `category` (exact and case-sensitive, at most 200 characters, stripped; blank
   means none) and `tags` (a comma list, blanks dropped, order-free, AND, at most 30
   tags of at most 200 characters) admit linked images only, through the linked
diff --git a/database/operations.py b/database/operations.py
index 385058a..2c81a71 100644
--- a/database/operations.py
+++ b/database/operations.py
@@ -213,15 +213,15 @@ JOIN prompts p ON p.id = COALESCE(pp.positive_prompt_id, gi.prompt_id)
 WHERE gi.file_path IS NOT NULL AND CASE WHEN {admit} THEN NOT EXISTS (
     SELECT 1 FROM generated_images other
     WHERE other.file_path = gi.file_path AND other.id <> gi.id) ELSE 0 END"""
-# The positive's text, its Library title or one of its tag names, or the pair
-# title, contains :pattern (LIKE, so ASCII case is ignored as in Library search).
-_LINK_FILTER_HIT = """(p.id IN (
+# The positive's text, its Library title or one of its tag names contains
+# :pattern (LIKE, so ASCII case is ignored as in Library search). A pair's own
+# title names its negative variant and is never searched (#249 D12).
+_LINK_FILTER_HIT = """p.id IN (
     SELECT hp.id FROM prompts hp
     LEFT JOIN prompt_library_state hls ON hls.prompt_id = hp.id
     WHERE hp.text LIKE :pattern ESCAPE '\\' OR hls.title LIKE :pattern ESCAPE '\\'
     OR hp.id IN (SELECT ht.prompt_id FROM prompt_tags ht JOIN tags hn
-                 ON hn.id = ht.tag_id WHERE hn.name LIKE :pattern ESCAPE '\\'))
-  OR pp.title LIKE :pattern ESCAPE '\\')"""
+                 ON hn.id = ht.tag_id WHERE hn.name LIKE :pattern ESCAPE '\\'))"""
 _LINK_FILTER_TAG = (
     "p.id IN (SELECT ft.prompt_id FROM prompt_tags ft JOIN tags fn "
     "ON fn.id = ft.tag_id WHERE fn.name = :tag{index})"
@@ -2340,7 +2340,8 @@ class PromptDatabase(
 
         Rows are limited to images whose linked positive has *category* and
         every tag in *tags* (exact names). ``hit`` says whether *q* occurs in
-        the positive's text, Library title or a tag name, or in the pair title.
+        the positive's text, Library title or a tag name; a pair's own title
+        names its negative variant and is not searched.
         Without category or tags only hits are returned, unless *every_link*
         asks for every linked image (the rating sort and the positive count).
         The positive id is the card's linked positive. One SELECT whose rows
diff --git a/tests/test_api_workspace_images.py b/tests/test_api_workspace_images.py
index 96d770a..131588e 100644
--- a/tests/test_api_workspace_images.py
+++ b/tests/test_api_workspace_images.py
@@ -1068,8 +1068,9 @@ class TestWorkspaceImages(ImageAPITestCase):
         cases = (
             # Six unambiguous positives; the ambiguous file and unlinked count none.
             ({}, 10, 6),
-            # Text, title, tag, pair title (its positive) and the filename match.
-            ({"q": "ZEBRA"}, 6, 4),
+            # Text, stored title, tag and the filename match (never a pair's
+            # own variant title, D12).
+            ({"q": "ZEBRA"}, 5, 3),
             ({"category": "Night"}, 4, 3),
             ({"tags": "a,b"}, 3, 2),
             ({"q": "plain_unlinked"}, 1, 0),
@@ -1200,18 +1201,24 @@ class TestWorkspaceImages(ImageAPITestCase):
 
         files = self._filter_fixtures()
         _, data = await self.get(q="ZEBRA", sort="name")
-        # Text, stored title, tag, pair title and filename; never an ambiguous link.
+        # Text, stored title, tag and filename; never an ambiguous link.
         self.assertEqual(
             self._names(data),
             [
                 "day_ab.png",
                 "night_ab.png",
-                "paired.png",
                 "titled.png",
                 "zebra_unlinked.png",
             ],
         )
-        self.assertEqual(data["total"], 5)
+        self.assertEqual(data["total"], 4)
+        # D12: a pair's own title names its negative variant, so the variant
+        # name alone matches nothing; the linked positive's text still does.
+        _, data = await self.get(q="Zebra pair")
+        self.assertEqual(self._names(data), [])
+        self.assertEqual(data["total"], 0)
+        _, data = await self.get(q="pair positive")
+        self.assertEqual(self._names(data), ["paired.png"])
         _, data = await self.get(q="0% w")
         self.assertEqual(self._names(data), ["percent.png"])
         _, data = await self.get(q="%")
@@ -1222,7 +1229,7 @@ class TestWorkspaceImages(ImageAPITestCase):
             self.api.db.get_image_paths_matching("zebra", None, []),
             {
                 normalize_image_path(str(files[name]))
-                for name in ("night_ab", "day_ab", "titled", "paired")
+                for name in ("night_ab", "day_ab", "titled")
             },
         )