It is generally accepted in the library world that an automated catalog means more accessible data for patrons, greater productivity for librarians, and an improvement in the sharing of bibliographic data among libraries. While the desirability of automation is not a controversial issue, some aspects of automating remain problematic. This article surveys the automation options available to libraries with Hebrew and Arabic script collections. It also examines the automation decisions that different libraries worldwide have made about automating such collections, particularly considering how their choices relate to overall prioritization and needs assessment at the institution. A library may choose to Romanize the cataloging data, to use nonRoman script cataloging, or to implement combinations of both. Standards and case studies are provided for each. Appendices contain sample machine-readable catalog records. (Contains 140 references.) (Author/BEW)