The XeTeX extended TeX engine provides a wealth of sophisticated features, and meets many of the long-felt needs of people working with multilingual or multi-script texts. I shall describe the use of XeLaTeX for typesetting Sanskrit, with both Roman- and Devanagari-script inputs, and Roman- and Devanagari-script outputs. I shall describe the complexities of getting differently hyphenated Sanskrit in different scripts. Finally, I shall offer an example of a free IBM XML tool that uses a XeLaTeX TEC file to auto-convert Sanskrit between Roman and Devanagari for screen display via HTML. If all this sounds a bit messy, it is. But the results are sometimes quite amazing, and open up exciting possibilities for the beautiful printing of Indian texts.