import pandas path = "test.csv" pandas.DataFrame({"A": [0,1], "B": [2,3]}).to_csv( path, encoding="utf-16") for encoding in ["utf-16","utf_16","UTF_16","UTF-16"]: try: pandas.io.parsers.read_csv( path, engine='c', encoding=encoding) print(encoding, "succeeded in c-pandas") except UnicodeDecodeError: print(encoding, "failed in c-pandas") try: pandas.io.parsers.read_csv( path, engine='python', encoding=encoding) print(encoding, "succeeded in pandas") except UnicodeDecodeError: print(encoding, "failed in pandas") try: with open(path, encoding=encoding) as file: file.read() print(encoding, "succeeded in open") except UnicodeDecodeError: print(encoding, "failed in open") 

utf-16 succeeded in c-pandas utf-16 succeeded in pandas utf-16 succeeded in open utf_16 succeeded in c-pandas utf_16 succeeded in pandas utf_16 succeeded in open UTF_16 succeeded in c-pandas UTF_16 succeeded in pandas UTF_16 succeeded in open UTF-16 succeeded in c-pandas UTF-16 succeeded in pandas UTF-16 succeeded in open 

utf-16 succeeded in c-pandas utf-16 succeeded in pandas utf-16 succeeded in open utf_16 failed in c-pandas utf_16 succeeded in pandas utf_16 succeeded in open UTF_16 failed in c-pandas UTF_16 succeeded in pandas UTF_16 succeeded in open UTF-16 failed in c-pandas UTF-16 succeeded in pandas UTF-16 succeeded in open