The paper introduces the grid file, a dynamic and symmetric multikey file structure designed to efficiently handle multikey access to records. Traditional file structures, such as inverted files, are extensions of single-key structures and suffer from deficiencies when dealing with highly dynamic files. The grid file addresses these issues by treating all keys symmetrically and adapting to the content of the file through dynamic partitioning and directory management. The authors start by discussing the limitations of single-key file structures and the challenges of designing balanced multikey structures. They propose a bitmap approach to represent the search space and introduce the concepts of grid partitions and grid directories. The grid file aims to achieve an upper bound of two disk accesses for single record retrieval and efficient handling of range and partially specified queries. The paper details the design decisions behind the grid file, including the choice of splitting and merging policies, implementation of the grid directory, and considerations for concurrent access. Simulation results are presented to evaluate the performance of the grid file, focusing on average bucket occupancy, directory size, and the geometry of bucket regions. The grid file is designed to be adaptable and efficient, with the ability to dynamically adjust its structure to the content of the file. It is particularly useful for applications requiring efficient multikey access, such as transaction-oriented systems and geometric processing. The paper concludes by discussing extensions to the grid file, including dynamic weighting of attributes and concurrent access.The paper introduces the grid file, a dynamic and symmetric multikey file structure designed to efficiently handle multikey access to records. Traditional file structures, such as inverted files, are extensions of single-key structures and suffer from deficiencies when dealing with highly dynamic files. The grid file addresses these issues by treating all keys symmetrically and adapting to the content of the file through dynamic partitioning and directory management. The authors start by discussing the limitations of single-key file structures and the challenges of designing balanced multikey structures. They propose a bitmap approach to represent the search space and introduce the concepts of grid partitions and grid directories. The grid file aims to achieve an upper bound of two disk accesses for single record retrieval and efficient handling of range and partially specified queries. The paper details the design decisions behind the grid file, including the choice of splitting and merging policies, implementation of the grid directory, and considerations for concurrent access. Simulation results are presented to evaluate the performance of the grid file, focusing on average bucket occupancy, directory size, and the geometry of bucket regions. The grid file is designed to be adaptable and efficient, with the ability to dynamically adjust its structure to the content of the file. It is particularly useful for applications requiring efficient multikey access, such as transaction-oriented systems and geometric processing. The paper concludes by discussing extensions to the grid file, including dynamic weighting of attributes and concurrent access.