Texts that include names or terms that use special characters with diacritics can present difficulties for indexers, and more so when the index is embedded in Microsoft Word or another program. Texts about Islam or the Middle East present several specific difficulties arising from the transliteration of Arabic into English. Even when diacritics are not used, many authors will still represent the letters ayn and hamza that, along with the definite article al-, should be ignored in the sort order. Indexing software can handle these tasks. However, when an index is to be embedded, the indexer must also try to get the indexing program to communicate with two or three software programs that code differently: the embedding program, the program in which the text is written, and possibly a layout program that will be used later. This article explains how to use Unicode characters and forced sorting in SKY Index 7 and SKY Index 8 for an index to be embedded in a Microsoft Word document using the DEXembed add-on. It touches on other issues to be considered if the document is to be flowed into InDesign, a layout program. This information is also relevant for other indexing programs and the transliteration of languages other than Arabic.