diff --git a/core/src/main/scala/org/apache/spark/serializer/KryoSerializer.scala b/core/src/main/scala/org/apache/spark/serializer/KryoSerializer.scala
index ea7af01121a1b..65f0ffcbe18ea 100644
--- a/core/src/main/scala/org/apache/spark/serializer/KryoSerializer.scala
+++ b/core/src/main/scala/org/apache/spark/serializer/KryoSerializer.scala
@@ -47,7 +47,7 @@ import org.apache.spark.internal.io.FileCommitProtocol._
 import org.apache.spark.network.util.ByteUnit
 import org.apache.spark.scheduler.{CompressedMapStatus, HighlyCompressedMapStatus}
 import org.apache.spark.storage._
-import org.apache.spark.unsafe.types.UTF8String
+import org.apache.spark.unsafe.types.{TimestampNanosVal, UTF8String}
 import org.apache.spark.util.{BoundedPriorityQueue, ByteBufferInputStream, NextIterator, SerializableConfiguration, SerializableJobConf, Utils}
 import org.apache.spark.util.collection.{BitSet, CompactBuffer}
 import org.apache.spark.util.io.ChunkedByteBuffer
@@ -236,6 +236,9 @@ class KryoSerializer(conf: SparkConf)
     kryo.register(classOf[ArrayBuffer[Any]])
     kryo.register(classOf[Array[Array[Byte]]])
     kryo.register(classOf[UTF8String])
+    // Nanosecond-timestamp min/max bounds inside cached-batch statistics rows
+    // (TimestampNanosColumnStats and the Arrow cache's vector-side stats).
+    kryo.register(classOf[TimestampNanosVal])
 
     // We can't load those class directly in order to avoid unnecessary jar dependencies.
     // We load them safely, ignore it if the class not found.
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/CacheTableInKryoSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/CacheTableInKryoSuite.scala
index 72a2da160547a..feee2bee298f3 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/CacheTableInKryoSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/CacheTableInKryoSuite.scala
@@ -57,6 +57,14 @@ class CacheTableInKryoSuite extends SharedSparkSession {
     }
   }
 
+  test("SPARK-58408: TimestampNanosVal stats should be registered in KryoSerializer") {
+    // Cached-batch statistics rows carry TimestampNanosVal min/max bounds for
+    // nanosecond-timestamp columns (TimestampNanosColumnStats), so persisting such a cache
+    // through Kryo with registrationRequired must find the class registered.
+    val df = sql("SELECT CAST('2025-01-06 12:30:45.123456789' AS TIMESTAMP_NTZ(9)) AS ts")
+    assert(df.persist(StorageLevel.DISK_ONLY).count() === 1)
+  }
+
   test("SPARK-51813 DefaultCachedBatchKryoSerializer do not propagate nulls") {
     val ks = new KryoSerializer(this.sparkConf)
     val kryo = ks.newKryo()