diff --git a/parquet-hadoop/README.md b/parquet-hadoop/README.md
index e40c93c047..b0c3443c7e 100644
--- a/parquet-hadoop/README.md
+++ b/parquet-hadoop/README.md
@@ -340,7 +340,7 @@ ParquetInputFormat to materialize records. It should be a the descendant class o
 
 **Property:** `parquet.compression.codec.zstd.bufferPool.enabled`  
 **Description:** If it is true, [RecyclingBufferPool](https://github.com/luben/zstd-jni/blob/master/src/main/java/com/github/luben/zstd/RecyclingBufferPool.java) is used.  
-**Default value:** `false`
+**Default value:** `true`
 
 ---
 
diff --git a/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/codec/ZstandardCodec.java b/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/codec/ZstandardCodec.java
index 73908da91e..fe4804acf4 100644
--- a/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/codec/ZstandardCodec.java
+++ b/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/codec/ZstandardCodec.java
@@ -47,7 +47,7 @@
 public class ZstandardCodec implements Configurable, CompressionCodec {
 
   public final static String PARQUET_COMPRESS_ZSTD_BUFFERPOOL_ENABLED = "parquet.compression.codec.zstd.bufferPool.enabled";
-  public final static boolean DEFAULT_PARQUET_COMPRESS_ZSTD_BUFFERPOOL_ENABLED = false;
+  public final static boolean DEFAULT_PARQUET_COMPRESS_ZSTD_BUFFERPOOL_ENABLED = true;
   public final static String PARQUET_COMPRESS_ZSTD_LEVEL = "parquet.compression.codec.zstd.level";
   public final static int DEFAULT_PARQUET_COMPRESS_ZSTD_LEVEL = 3;
   public final static String PARQUET_COMPRESS_ZSTD_WORKERS = "parquet.compression.codec.zstd.workers";