The costs of creating and maintaining permanent files and retrieving keyword title indexes in TRIAL (Technique for Retrieving Information from Abstracts of Literature) are presented. This system, which is used for bibliographic control and file efficiency, is relatively easy to learn and use, and popular with school, academic, business, and governmental personnel. Large data bases are anticipated, and the costs for building 1000, 3000, 5000, 7000, and 9000 record data bases are reported here based on a model devised to show how most users use TRIAL. Results show that costs are linear in relation to the number of records needed, and that each record costs approximately seven and one-half cents to process. (Author/MBR)