Nestled underneath the tentorium in the posterior cranial fossa, the cerebellum is a remarkable brain structure. It plays an essential role in motor coordination [1] and cognitive function [2]. The cerebellum is divided into two hemispheres and a midline zone called the vermis. Further subdivision is based on hierarchical folds known as lobes (anterior, superior posterior, inferior posterior, and flocculonodular) and lobules (identified by Roman numerals I–X). Brodmann areas have been used to subdivide the cerebral cortex for more than a century [3]. However, unlike the cerebrum, the cerebellum has a rather homogeneous cytoarchitecture, and although some evidence is increasingly challenging this assumption [4], the cerebellar lobules are still the preferred way to subdivide the structure. Different groups have advanced the application of computational techniques to segment the cerebellum with the objective of developing a fully automatic algorithm. From atlas-based approaches to the newest deep learning models, each novel software alternative becomes more accurate and requires less human intervention. In a data-driven world, such methods are essential for advancing our understanding of the cerebellum and its role in brain function and dysfunction. It is worth noting that different methods have their advantages and limitations, requiring investigators to choose wisely to maximize the possible insights of their research.