diff --git a/apps/common/parsers/csv.py b/apps/common/parsers/csv.py
index 4cd626cef..d96e9d2ef 100644
--- a/apps/common/parsers/csv.py
+++ b/apps/common/parsers/csv.py
@@ -3,6 +3,7 @@
 
 import json
 import chardet
+import codecs
 import unicodecsv
 
 from rest_framework.parsers import BaseParser
@@ -82,6 +83,7 @@ class JMSCSVParser(BaseParser):
 
         try:
             stream_data = stream.read()
+            stream_data = stream_data.strip(codecs.BOM_UTF8)
             detect_result = chardet.detect(stream_data)
             encoding = detect_result.get("encoding", "utf-8")
             binary = self._universal_newlines(stream_data)
diff --git a/apps/common/renders/csv.py b/apps/common/renders/csv.py
index 7e57a97e1..f917edd4c 100644
--- a/apps/common/renders/csv.py
+++ b/apps/common/renders/csv.py
@@ -2,6 +2,7 @@
 #
 
 import unicodecsv
+import codecs
 from datetime import datetime
 
 from six import BytesIO
@@ -73,7 +74,8 @@ class JMSCSVRender(BaseRenderer):
             table = self._gen_table(data, header, labels)
 
             csv_buffer = BytesIO()
-            csv_writer = unicodecsv.writer(csv_buffer, encoding='utf-8-sig')
+            csv_buffer.write(codecs.BOM_UTF8)
+            csv_writer = unicodecsv.writer(csv_buffer, encoding='utf-8')
             for row in table:
                 csv_writer.writerow(row)