diff --git a/csvs_to_sqlite/cli.py b/csvs_to_sqlite/cli.py
index 5cca320..259b288 100644
--- a/csvs_to_sqlite/cli.py
+++ b/csvs_to_sqlite/cli.py
@@ -116,6 +116,12 @@
     is_flag=True,
     help="Skip adding full-text index on values extracted using --extract-column (default is to add them)",
 )
+@click.option(
+    "--no-na-filter",
+    "no_na_filter",
+    is_flag=True,
+    help="Skip detection of missing value markers (empty strings and the value of na_values). See pandas.read_csv() documentation",
+)
 @click.version_option()
 def cli(
     paths,
@@ -136,6 +142,7 @@ def cli(
     filename_column,
     no_index_fks,
     no_fulltext_fks,
+    no_na_filter
 ):
     """
     PATHS: paths to individual .csv files or to directories containing .csvs
@@ -162,7 +169,7 @@ def cli(
     sql_type_overrides = None
     for name, path in csvs.items():
         try:
-            df = load_csv(path, separator, skip_errors, quoting, shape)
+            df = load_csv(path, separator, skip_errors, quoting, shape, no_na_filter)
             df.table_name = table or name
             if filename_column:
                 df[filename_column] = name
diff --git a/csvs_to_sqlite/utils.py b/csvs_to_sqlite/utils.py
index bb4ae4f..bfa13e8 100644
--- a/csvs_to_sqlite/utils.py
+++ b/csvs_to_sqlite/utils.py
@@ -25,6 +25,7 @@ def load_csv(
     skip_errors,
     quoting,
     shape,
+    no_na_filter,
     encodings_to_try=("utf8", "latin-1"),
 ):
     usecols = None
@@ -41,6 +42,7 @@ def load_csv(
                     low_memory=True,
                     encoding=encoding,
                     usecols=usecols,
+                    na_filter=not no_na_filter
                 )
             except UnicodeDecodeError:
                 continue