Thanks for everyone's reponses to this. It sounds like a couple of things have to happen to ensure that templates work for everyone. 1. Template editors like me need to be sure their text editors will save documents with their original encodeing. 2. HTML in Koha templates must reflect the encoding of that particular page. Is there any reason not to go with utf-8 for all templates as MJR/slef suggests? That is to say, is there any situation with Koha where utf-8 would be inappropriate? -- Owen