diff --git a/src/sql/das/ob_das_scan_op.cpp b/src/sql/das/ob_das_scan_op.cpp
index 1d7ba10e6b..b65b182383 100644
--- a/src/sql/das/ob_das_scan_op.cpp
+++ b/src/sql/das/ob_das_scan_op.cpp
@@ -313,6 +313,7 @@ int ObDASScanOp::init_scan_param()
         }
       }
       scan_param_.external_file_format_.csv_format_.file_column_nums_ = static_cast<int64_t>(max_idx);
+      scan_param_.external_file_format_.csv_format_.ignore_extra_fields_ = true;
     }
   }
   LOG_DEBUG("init scan param", K(ret), K(scan_param_));
diff --git a/src/sql/engine/cmd/ob_load_data_parser.h b/src/sql/engine/cmd/ob_load_data_parser.h
index fd43ec831a..14870e89db 100644
--- a/src/sql/engine/cmd/ob_load_data_parser.h
+++ b/src/sql/engine/cmd/ob_load_data_parser.h
@@ -40,6 +40,7 @@ struct ObCSVGeneralFormat {
     cs_type_(common::CHARSET_INVALID),
     skip_header_lines_(0),
     skip_blank_lines_(false),
+    ignore_extra_fields_(false),
     trim_space_(false),
     null_if_(),
     empty_field_as_null_(false),
@@ -65,6 +66,7 @@ struct ObCSVGeneralFormat {
   common::ObCharsetType cs_type_; // charset type of format strings
   int64_t skip_header_lines_;
   bool skip_blank_lines_;
+  bool ignore_extra_fields_;
   bool trim_space_;
   common::ObArrayWrap<common::ObString> null_if_;
   bool empty_field_as_null_;
@@ -442,7 +444,8 @@ int ObCSVGeneralParser::scan_proto(const char *&str,
     }
     if (OB_LIKELY(find_new_line) || is_end_file) {
       if (!format_.skip_blank_lines_ || field_idx > 0) {
-        if (field_idx != format_.file_column_nums_) {
+        if (field_idx < format_.file_column_nums_
+            || (field_idx > format_.file_column_nums_ && !format_.ignore_extra_fields_)) {
           ret = handle_irregular_line(field_idx, line_no, errors);
         }
         if (OB_SUCC(ret)) {