I am using XStream 1.0.2. The following two codes produce the same output: --------------- Sample 1 -------------- final XStream x1 = new XStream(new DomDriver("UTF-8")); x1.toXML(teststring, new FileWriter("test-encoding-1.xml")); --------------- Sample 2 -------------- final XStream x2 = new XStream(new DomDriver()); x2.toXML(teststring, new FileWriter("test-encoding-2.xml")); --------------- End Samples ----------- Given the teststring variable is "Ää Öö Üu", the output for both is identical: 00000000h: 3C 73 74 72 69 6E 67 3E ; <string> 00000008h: C4 E4 20 D6 F6 20 DC FC ; Ää Öö Üü 00000010h: 3C 2F 73 74 72 69 6E 67 ; </string 00000018h: 3E ; > Obviously, the umlaut encoding does not take place in UTF-8, which (to my expectations) would look like this: 00000000h: 3C 73 74 72 69 6E 67 3E ; <string> 00000008h: C3 84 C3 A4 20 C3 96 C3 ; Ää Ã-à 00000010h: B6 20 C3 9C C3 BC 3C 2F ; ¶ Üü</ 00000018h: 73 74 72 69 6E 67 3E ; string> The encoding of the above code snippets rather take place in my Windows XPs default encoding schema ISO-8859-1.