The article provides an overview of multidimensional access methods used in spatial databases, which are essential for efficient search operations such as point queries and region queries. The authors, Volker Gaede and Oliver Günther, discuss the challenges and requirements of spatial data management, including the dynamic nature of spatial data, the need for secondary/tertiary storage management, and the lack of a standard spatial algebra or query language. They introduce various data structures and access methods, categorizing them into point access methods (PAMs) and spatial access methods (SAMs). PAMs are designed for point databases, while SAMs handle extended objects like rectangles or polyhedra. The article covers classical one-dimensional access methods like linear hashing, extendible hashing, and B-trees, as well as multidimensional variants such as k-d-trees, BSP-trees, BD-trees, and quad trees. It also discusses hybrid approaches and their performance characteristics, emphasizing the importance of secondary storage management and the trade-offs between time and space efficiency. The article concludes with a discussion on theoretical and experimental results comparing the performance of different access methods.The article provides an overview of multidimensional access methods used in spatial databases, which are essential for efficient search operations such as point queries and region queries. The authors, Volker Gaede and Oliver Günther, discuss the challenges and requirements of spatial data management, including the dynamic nature of spatial data, the need for secondary/tertiary storage management, and the lack of a standard spatial algebra or query language. They introduce various data structures and access methods, categorizing them into point access methods (PAMs) and spatial access methods (SAMs). PAMs are designed for point databases, while SAMs handle extended objects like rectangles or polyhedra. The article covers classical one-dimensional access methods like linear hashing, extendible hashing, and B-trees, as well as multidimensional variants such as k-d-trees, BSP-trees, BD-trees, and quad trees. It also discusses hybrid approaches and their performance characteristics, emphasizing the importance of secondary storage management and the trade-offs between time and space efficiency. The article concludes with a discussion on theoretical and experimental results comparing the performance of different access methods.