An analysis of existing computerized data banks in science and technology reveals that nearly half of them involve the storage and retrieval of bibliographic data. Activity in this area has been independent and autonomous. This situation is now giving way to a new environment which involves cooperation, standards, and a rigorous rational analysis of the traditional raw materials and processes of librarianship. There is a genuine rapprochement between librarians and computer specialists, resulting in a scientific approach to the problems posed by the control and retrieval of bibliographic entities. It is observed that the design of systems dealing with bibliographic data must respond to pressures generated by the structure of the data itself, as well as to inter-system and user requirements. These pressures can take effect on several levels, ranging from programming subroutines, to file structure, to the organization of access modes and output formats. Examples of each are provided. This is a critical period for library automation, for the standardization of machine input records, and for the design of retrieval systems dealing with these records. Decisions made today will affect the access to bibliographic data for decades to come. (Author/JB)