Chemometric analysis of complex systems often involves large datasets. Efficiently managing these datasets requires careful sample subset selection, encompassing two key tasks: selecting a representative sample subset for initial analysis and partitioning data for model calibration and validation. This review provides a comprehensive overview of 28 sample subset selection methods developed within the chemometrics field. For the first time, we classify these methods into seven distinct categories based on their underlying principles: sampling-based, distance-based, clustering-inspired, experimental design-inspired, variable selection-inspired, outlier detection-inspired, and preprocessing-inspired approaches. We systematically discuss the principles, advantages, disadvantages, and typical applications of each method. This consolidation serves as a valuable resource for researchers, facilitating the informed selection of appropriate sample subset selection strategies prior to multivariate calibration or chemical pattern recognition tasks.