High-entropy alloys (HEAs) have attracted significant attention because of their exceptional mechanical properties and potential for discovering new compositions. However, owing to their complex chemical makeup, understanding the underlying physical mechanisms and designing novel alloys through traditional theoretical and computational methods are challenging. Machine learning (ML) offers a promising solution to address these challenges. This review begins with a discussion of the critical data preprocessing techniques required for effective ML applications in the HEA domain. Subsequently, this review discusses various ML models, including supervised and unsupervised algorithms, focusing on guiding readers through algorithm selection and model evaluation. These fundamental aspects are essential for understanding the complexities of applying ML in HEA research. Further, this review highlights various successful applications of ML in HEA research. These include optimisation of the alloy composition, processing parameters, and microstructural characteristics to enhance the mechanical properties. In addition, this review examines the use of ML for performance-driven reverse engineering of HEA compositions, enabling the rapid identification of new high-performance alloy designs. The novelty of this review is its comprehensive and integrative approach to ML applications in HEAs. Unlike previous studies that focused on specific ML techniques or isolated use cases, this review explores the transformative potential of ML across the entire HEA research landscape. By covering the full spectrum from fundamental data preprocessing and model selection to a diverse range of practical applications, this review provides insights into how ML can accelerate the discovery and development of high-performance HEAs. This multifaceted perspective covers the synergistic interplay between various ML methodologies and their impact on HEA research, opening up new opportunities for innovation that may not have been fully explored in more specialized studies. Considering the extensive studies on ML discussed in this review, it can be concluded that ML has revolutionised the design and development of HEAs. By optimising alloy composition, processing parameters, and microstructural characteristics, ML-driven approaches have unlocked the potential for engineering high-performance HEAs with tailored mechanical properties. Moreover, the use of ML in performance-driven reverse engineering has enabled the rapid identification of promising HEA compositions, thereby accelerating the discovery of novel materials.