diff --git a/src/test/regress/expected/join.out b/src/test/regress/expected/join.out
index 41521f275c8..d2a3adeb6d8 100644
--- a/src/test/regress/expected/join.out
+++ b/src/test/regress/expected/join.out
@@ -10060,3 +10060,38 @@ SELECT COUNT(*) FROM onek t1 LEFT JOIN tenk1 t2
  19000
 (1 row)
 
+--
+-- Test that estimate_hash_bucket_stats correctly handles skewed data
+-- with restriction clauses.
+--
+create table hjbucket (val int not null, filt int not null);
+insert into hjbucket
+select
+    case when g <= 200 then 0 else ((g - 201) % 100) + 1 end,
+    g % 10
+from generate_series(1, 2000) g;
+create index on hjbucket(filt);
+analyze hjbucket;
+set work_mem to '64kB';
+set enable_nestloop to off;
+set enable_mergejoin to off;
+explain (costs off)
+select * from hjbucket h1
+  join hjbucket h2 on h1.val = h2.val
+  where h1.filt = 5;
+                        QUERY PLAN                        
+----------------------------------------------------------
+ Hash Join
+   Hash Cond: (h2.val = h1.val)
+   ->  Seq Scan on hjbucket h2
+   ->  Hash
+         ->  Bitmap Heap Scan on hjbucket h1
+               Recheck Cond: (filt = 5)
+               ->  Bitmap Index Scan on hjbucket_filt_idx
+                     Index Cond: (filt = 5)
+(8 rows)
+
+reset work_mem;
+reset enable_nestloop;
+reset enable_mergejoin;
+drop table hjbucket;
diff --git a/src/test/regress/sql/join.sql b/src/test/regress/sql/join.sql
index 4acd2512004..abcbf13d379 100644
--- a/src/test/regress/sql/join.sql
+++ b/src/test/regress/sql/join.sql
@@ -3841,3 +3841,31 @@ SELECT COUNT(*) FROM onek t1 LEFT JOIN tenk1 t2
     ON (t2.thousand = t1.tenthous OR t2.thousand = t1.thousand);
 SELECT COUNT(*) FROM onek t1 LEFT JOIN tenk1 t2
     ON (t2.thousand = t1.tenthous OR t2.thousand = t1.thousand);
+
+--
+-- Test that estimate_hash_bucket_stats correctly handles skewed data
+-- with restriction clauses.
+--
+create table hjbucket (val int not null, filt int not null);
+insert into hjbucket
+select
+    case when g <= 200 then 0 else ((g - 201) % 100) + 1 end,
+    g % 10
+from generate_series(1, 2000) g;
+create index on hjbucket(filt);
+analyze hjbucket;
+
+set work_mem to '64kB';
+set enable_nestloop to off;
+set enable_mergejoin to off;
+
+explain (costs off)
+select * from hjbucket h1
+  join hjbucket h2 on h1.val = h2.val
+  where h1.filt = 5;
+
+reset work_mem;
+reset enable_nestloop;
+reset enable_mergejoin;
+
+drop table hjbucket;