STARTING FROM A MODEL OF CONTEMPORARY MECHANIZED RETRIEVAL SYSTEMS AND THE CHARACTERISTICS OF INDEXING LANGUAGES USED THEREIN, THE AUTHORS DEVELOP A RATIONAL BASIS FOR USE OF THE UNIVERSAL DECIMAL CLASSIFICATION (UDC) IN THIS CONTEXT. PRACTICAL DESIGN CONSIDERATIONS FOR THE USE OF UDC IN A MECHANIZED RETRIEVAL SYSTEM ARE DISCUSSED. EXAMPLES ARE REPORTED OF THE USE OF UDC AS THE INDEXING LANGUAGE WITH THE COMBINED FILE SEARCH SYSTEM, AN EXISTING RETRIEVAL SYSTEM FOR THE IBM 1401, USED BY SEVERAL LARGE INFORMATION CENTERS IN THE UNITED STATES. FINALLY, THE AUTHORS DISCUSS HOW UDC MIGHT BE USED AS A QUERY LANGUAGE IN A TYPICAL RETRIEVAL SYSTEM OF THE NEAR FUTURE IN WHICH THE USER INTERACTS DIRECTLY WITH THE COMPUTER-STORED DOCUMENT REFERENCE FILE. THE AUTHORS CONCLUDE THAT IT IS TECHNICALLY FEASIBLE TO USE UDC IN MECHANIZED RETRIEVAL SYSTEMS AND THAT, UNDER CERTAIN CONDITIONS, IT MAY BE DESIRABLE. SOME OF THESE CONDITIONS ARE THE EXISTENCE OF LARGE FILES ALREADY INDEXED BY UDC, STAFF ALREADY TRAINED FOR ITS USE, AND EXTENSIVE INTERNATIONAL USE OR EXCHANGE OF MATERIALS OF THE SYSTEM. THIS PAPER WAS PRESENTED AT THE FIC/IFIP CONFERENCE ON MECHANIZED INFORMATION STORAGE, RETRIEVAL, AND DISSEMINATION (ROME, JUNE 15, 1967). THIS DOCUMENT IS AVAILABLE AS PB-176-152 FROM THE CLEARINGHOUSE FOR FEDERAL SCIENTIFIC AND TECHNICAL INFORMATION, SPRINGFIELD, VIRGINIA 22151, $3.00 FOR HARD COPY, $0.65 FOR MICROFICHE, 37 PAGES. (AUTHOR)