diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/encoders/ExpressionEncoder.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/encoders/ExpressionEncoder.scala
index 74d7a5e7a675..654f39393636 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/encoders/ExpressionEncoder.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/encoders/ExpressionEncoder.scala
@@ -325,11 +325,19 @@ case class ExpressionEncoder[T](
   assert(serializer.forall(_.references.isEmpty), "serializer cannot reference any attributes.")
   assert(serializer.flatMap { ser =>
     val boundRefs = ser.collect { case b: BoundReference => b }
-    assert(boundRefs.nonEmpty,
-      "each serializer expression should contain at least one `BoundReference`")
+    assert(boundRefs.nonEmpty || isEmptyStruct(ser),
+      "each serializer expression should contain at least one `BoundReference` or it " +
+      "should be an empty struct. This is required to ensure that there is a reference point " +
+      "for the serialized object or that the serialized object is intentionally left empty."
+    )
     boundRefs
   }.distinct.length <= 1, "all serializer expressions must use the same BoundReference.")
 
+  private def isEmptyStruct(expr: NamedExpression): Boolean = expr.dataType match {
+    case struct: StructType => struct.isEmpty
+    case _ => false
+  }
+
   /**
    * Returns a new copy of this encoder, where the `deserializer` is resolved and bound to the
    * given schema.
diff --git a/sql/core/src/test/resources/sql-tests/analyzer-results/selectExcept.sql.out b/sql/core/src/test/resources/sql-tests/analyzer-results/selectExcept.sql.out
index 8643d40b886b..48c7ad10ed78 100644
--- a/sql/core/src/test/resources/sql-tests/analyzer-results/selectExcept.sql.out
+++ b/sql/core/src/test/resources/sql-tests/analyzer-results/selectExcept.sql.out
@@ -121,6 +121,18 @@ Project [id#x, name#x, named_struct(f1, data#x.f1, s2, named_struct(f3, data#x.s
                +- LocalRelation [id#x, name#x, data#x]
 
 
+-- !query
+SELECT * EXCEPT (data.f1, data.s2) FROM tbl_view
+-- !query analysis
+Project [id#x, name#x, named_struct() AS data#x]
++- SubqueryAlias tbl_view
+   +- View (`tbl_view`, [id#x,name#x,data#x])
+      +- Project [cast(id#x as int) AS id#x, cast(name#x as string) AS name#x, cast(data#x as struct<f1:int,s2:struct<f2:int,f3:string>>) AS data#x]
+         +- Project [id#x, name#x, data#x]
+            +- SubqueryAlias tbl_view
+               +- LocalRelation [id#x, name#x, data#x]
+
+
 -- !query
 SELECT * EXCEPT (id, name, data) FROM tbl_view
 -- !query analysis
diff --git a/sql/core/src/test/resources/sql-tests/inputs/selectExcept.sql b/sql/core/src/test/resources/sql-tests/inputs/selectExcept.sql
index e07e4f1117c2..08d56aeda0a8 100644
--- a/sql/core/src/test/resources/sql-tests/inputs/selectExcept.sql
+++ b/sql/core/src/test/resources/sql-tests/inputs/selectExcept.sql
@@ -20,6 +20,7 @@ SELECT * EXCEPT (data) FROM tbl_view;
 SELECT * EXCEPT (data.f1) FROM tbl_view;
 SELECT * EXCEPT (data.s2) FROM tbl_view;
 SELECT * EXCEPT (data.s2.f2) FROM tbl_view;
+SELECT * EXCEPT (data.f1, data.s2) FROM tbl_view;
 -- EXCEPT all columns
 SELECT * EXCEPT (id, name, data) FROM tbl_view;
 -- EXCEPT special character names
diff --git a/sql/core/src/test/resources/sql-tests/results/selectExcept.sql.out b/sql/core/src/test/resources/sql-tests/results/selectExcept.sql.out
index 6f6ba9097342..2621782342cc 100644
--- a/sql/core/src/test/resources/sql-tests/results/selectExcept.sql.out
+++ b/sql/core/src/test/resources/sql-tests/results/selectExcept.sql.out
@@ -121,6 +121,20 @@ struct<id:int,name:string,data:struct<f1:int,s2:struct<f3:string>>>
 70	name7	{"f1":7,"s2":{"f3":"g"}}
 
 
+-- !query
+SELECT * EXCEPT (data.f1, data.s2) FROM tbl_view
+-- !query schema
+struct<id:int,name:string,data:struct<>>
+-- !query output
+10	name1	{}
+20	name2	{}
+30	name3	{}
+40	name4	{}
+50	name5	{}
+60	name6	{}
+70	name7	{}
+
+
 -- !query
 SELECT * EXCEPT (id, name, data) FROM tbl_view
 -- !query schema