Molecular docking is a powerful tool in drug discovery, particularly for structure-based virtual screening. This review introduces the basic theories and methods of molecular docking, including sampling algorithms and scoring functions. It discusses the performance and differences among available docking software, highlighting the challenges of flexible receptor docking. A recently developed Local Move Monte Carlo (LMMC) approach is introduced as a potential solution for flexible receptor docking. The article also provides three application examples of molecular docking in drug discovery, demonstrating its effectiveness in identifying new active compounds and optimizing lead structures. The review emphasizes the importance of accurate and efficient scoring functions and the need for further improvements in docking methodologies to better predict binding affinities.Molecular docking is a powerful tool in drug discovery, particularly for structure-based virtual screening. This review introduces the basic theories and methods of molecular docking, including sampling algorithms and scoring functions. It discusses the performance and differences among available docking software, highlighting the challenges of flexible receptor docking. A recently developed Local Move Monte Carlo (LMMC) approach is introduced as a potential solution for flexible receptor docking. The article also provides three application examples of molecular docking in drug discovery, demonstrating its effectiveness in identifying new active compounds and optimizing lead structures. The review emphasizes the importance of accurate and efficient scoring functions and the need for further improvements in docking methodologies to better predict binding affinities.