diff --git a/.travis.yml b/.travis.yml
index 0293824..3be998f 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -26,7 +26,7 @@ script:
       < data/test-input.csv
     | docker-compose run psql_client
   - coverage run
-      -a --source=csv2sql -m csv2sql data tbl < data/test-input.csv
+      -a --source=csv2sql -m csv2sql data -r tbl < data/test-input.csv
     | docker-compose run psql_client
 after_success:
   - coveralls
diff --git a/csv2sql/main.py b/csv2sql/main.py
index 709b71b..06ee9b5 100644
--- a/csv2sql/main.py
+++ b/csv2sql/main.py
@@ -79,9 +79,11 @@ def _dump_schema(args, in_file=None):
     )
 
 
-def _dump_data(args, in_file=None):
+def _dump_data(args, in_file=None, rebuild=None):
     if not in_file:
         in_file = args.in_file
+    if rebuild is None:
+        rebuild = args.rebuild
 
     # Skip the header.
     reader = csv.reader(in_file, delimiter=args.delimiter)
@@ -92,6 +94,7 @@ def _dump_data(args, in_file=None):
         args.table_name,
         reader,
         args.null,
+        rebuild,
     )
 
 
@@ -100,7 +103,7 @@ def _dump_all(args):
         _dump_schema(args, in_file=file_iterator)
         file_iterator.rewind()
         frozen_file_iterator = file_iterator.freeze()
-        _dump_data(args, in_file=frozen_file_iterator)
+        _dump_data(args, in_file=frozen_file_iterator, rebuild=False)
 
 
 def parse_args(arguments):
@@ -146,7 +149,7 @@ def parse_args(arguments):
     schema_factory = argparse.ArgumentParser(add_help=False)
     schema_factory.add_argument(
         '-r', '--rebuild', action='store_true',
-        help='Rebuild the table by "DROP TABLE IF EXISTS".')
+        help='Rebuild the table by a query such as "DROP TABLE IF EXISTS".')
     schema_factory.add_argument(
         '--lines-for-inference', metavar='NUM',
         help=('Num lines to identify column types.'
@@ -154,6 +157,12 @@ def parse_args(arguments):
               ' used to identify them. [default: 0]'),
         type=int, default=0)
 
+    # insertion_factory.
+    insertion_factory = argparse.ArgumentParser(add_help=False)
+    insertion_factory.add_argument(
+        '-r', '--rebuild', action='store_true',
+        help='Rebuild the table by a query such as "TRUNCATE TABLE".')
+
     # pattern_readable.
     pattern_readable = argparse.ArgumentParser(add_help=False)
     pattern_readable.add_argument(
@@ -164,8 +173,9 @@ def parse_args(arguments):
     schema_dumper = [
         readable, writable, query_engine_dependent, csv_readable,
         query_factory, schema_factory, pattern_readable]
-    query_dumper = [
-        readable, writable, query_engine_dependent, csv_readable, query_factory]
+    insertion_dumper = [
+        readable, writable, query_engine_dependent, csv_readable,
+        query_factory, insertion_factory]
     pattern_dumper = [writable, query_engine_dependent, pattern_readable]
 
     # Main.
@@ -184,7 +194,7 @@ def parse_args(arguments):
         'schema', help='Schema queries.', parents=schema_dumper,
     ).set_defaults(command=_dump_schema)
     subparsers.add_parser(
-        'data', help='Data-insertion queries.', parents=query_dumper,
+        'data', help='Data-insertion queries.', parents=insertion_dumper,
     ).set_defaults(command=_dump_data)
     subparsers.add_parser(
         'pattern', help='Type-inference patterns.', parents=pattern_dumper,
diff --git a/csv2sql/queryengines/postgresql.py b/csv2sql/queryengines/postgresql.py
index e3f7a5b..c218215 100644
--- a/csv2sql/queryengines/postgresql.py
+++ b/csv2sql/queryengines/postgresql.py
@@ -98,8 +98,16 @@ def write_schema_statement(out_stream, table_name, column_types, rebuild=False):
     out_stream.write(_LINE_TERMINATOR)
 
 
-def write_insert_statement(out_stream, table_name, reader, null_value):
-    """Write the insert query into `out_stream`."""
+def write_insert_statement(
+        out_stream, table_name, reader, null_value, rebuild=False):
+    """Write the insert query into `out_stream`.
+    When `rebuild` is true, it prepends the query
+    'TRUNCATE TABLE `table_name`.
+    """
+    if rebuild:
+        out_stream.write('TRUNCATE TABLE {0};'.format(table_name))
+        out_stream.write(_LINE_TERMINATOR)
+
     out_stream.write(
         'COPY {0} FROM STDIN WITH NULL \'{1}\' CSV;'.format(
             table_name,
diff --git a/csv2sql/tests/test_main.py b/csv2sql/tests/test_main.py
index 46d7d77..535c235 100644
--- a/csv2sql/tests/test_main.py
+++ b/csv2sql/tests/test_main.py
@@ -28,7 +28,7 @@ def test_schema_dumper(self, command_name):
     @parameterized.expand([
         ('data',),
     ])
-    def test_query_dumper(self, command_name):
+    def test_insertion_dumper(self, command_name):
         arguments = [command_name, 'table-name']
         actual = parse_args(arguments)
         eq_(actual.table_name, 'table-name')
@@ -36,5 +36,6 @@ def test_query_dumper(self, command_name):
         ok_(hasattr(actual, 'out_file'))
         ok_(hasattr(actual, 'null'))
         ok_(hasattr(actual, 'delimiter'))
+        ok_(hasattr(actual, 'rebuild'))
         ok_(hasattr(actual, 'command'))
         ok_(hasattr(actual, 'query_engine'))