Romanization as a method in the bibliographic environment makes retrieval difficult in many aspects. The object of this thesis is to illustrate the importance and the possibility to include non-Roman script in the records of the library catalogue. The questions at issue are about the consequences of romanization, the technical requirements for implementing non-Roman script, and bibliographical resources that have been developed during the last 20 years. The thesis is based on a literature study and correspondence with people who are familiar with the subject. The thesis points out that Romanization causes information to be distorted in various ways and is inconsistent with the requirements of exactitude in bibliographic control. Different Romanization schemes make information exchange difficult and the method impede access to the literature for the readers. In the computerised environment different character-sets have been used in different settings to represent non-Roman script. The need for a single global character-set led to the development of the Unicode standard which has a capacity to encode all characters used for written languages. Queens Library in New York and SOAS in London are examples of libraries implementing original script. They use the bibliographic resources developed by OCLC and RLG which also are described in the thesis. In Sweden both LIBRIS and BURK are interested in the subject and discussions between the International Library of Stockholm and Btj/BURK are in progress. The conclusion is that Romanization constitutes a conflict between the treatment and the retrieval of document with non-Roman script. The solution is to describe the documents in their proper script and so treat all documents equivalently. This will increase retrieval and make the library more user-friendly.