Robert Vojta added a comment - 13/Jul/09 12:43 PM You can look at http://greenbytes.de/tech/tc2231/ for more info how this can be done. The problem is that there's no standard way how to do this for all browsers and you have to handle separate versions of IE as well I've workaround with the following code in Java and I pass function return value to OutputResource filename parameter. This works for IE / Firefox ... private static String encodeForIE(String fileName) throws UnsupportedEncodingException { /* * http://greenbytes.de/tech/tc2231/#attwithfnrawpctenca * * IE decodes %XY to characters and than if it detects * UTF-8 stream (after decoding of %XY), than it creates * UTF-8 string. * * We use this behavior to offer correct file name * for download. */ StringBuilder encodedFileName = new StringBuilder(); encodedFileName.append("\""); // ICEfaces 1.7.2 bug encodedFileName.append(URLEncoder.encode(fileName, "UTF-8").replaceAll("\\+", "%20")); encodedFileName.append("\""); // ICEfaces 1.7.2 bug return encodedFileName.toString(); } private static String encodeForFirefox(String fileName) throws UnsupportedEncodingException { /* http://greenbytes.de/tech/tc2231/#attwithutf8fnplain * Firefox is trying to be smart and decodes UTF-8 characters written as ISO-8859-1 bytes as UTF-8. * We use this Firefox behavior to offer correct file name for download. */ StringBuilder encodedFileName = new StringBuilder(); encodedFileName.append("\""); // ICEfaces 1.7.2 bug byte[] utf8Bytes = fileName.getBytes("UTF-8"); for (byte b : utf8Bytes) { char ch = (char) b; encodedFileName.append(ch); } encodedFileName.append("\""); // ICEfaces 1.7.2 bug return encodedFileName.toString(); }