import codecs doc = codecs.open('document','rU','UTF-16') #open for reading with "universal" type set df = pandas.read_csv(doc, sep=' ')