The SLINK algorithm, developed by R. Sibson, is an efficient method for performing single-link (nearest-neighbour) cluster analysis on dissimilarity coefficients. It achieves optimal storage and computational complexity, making it feasible for large datasets with up to $10^9$ objects. The algorithm uses a pointer representation to store the dendrogram, which can be efficiently updated as new objects are added. This representation can be converted into a tree-diagram for visualization. The SLINK algorithm is implemented in FORTRAN and is designed to be easily programmable. The paper also discusses the limitations of other clustering methods and the advantages of the single-link method, including its ability to handle large datasets and its robustness in the presence of distribution-mixture problems.The SLINK algorithm, developed by R. Sibson, is an efficient method for performing single-link (nearest-neighbour) cluster analysis on dissimilarity coefficients. It achieves optimal storage and computational complexity, making it feasible for large datasets with up to $10^9$ objects. The algorithm uses a pointer representation to store the dendrogram, which can be efficiently updated as new objects are added. This representation can be converted into a tree-diagram for visualization. The SLINK algorithm is implemented in FORTRAN and is designed to be easily programmable. The paper also discusses the limitations of other clustering methods and the advantages of the single-link method, including its ability to handle large datasets and its robustness in the presence of distribution-mixture problems.