Abstract: Groundwater is an important strategic resource for ensuring water security and maintaining ecosystem stability. Groundwater monitoring networks provide essential information for groundwater resource management and pollution prevention, and their design directly affects the representativeness of monitoring data and the reliability of management decisions. Traditional monitoring network design approaches based primarily on expert experience and technical specifications can no longer satisfy the increasing demand for refined groundwater management. Consequently, groundwater monitoring network optimization has become an important research topic in hydrogeology and environmental sciences. In recent years, statistics, geostatistics, optimization theory, numerical simulation, and artificial intelligence have provided a wide range of methods for groundwater monitoring network optimization. However, existing studies have mainly focused on individual algorithms or specific technical approaches, and a systematic methodological classification and comparison remain lacking. Based on a comprehensive review of recent studies, this paper establishes a methodological framework for groundwater monitoring network optimization according to information sources and theoretical foundations. Existing methods are classified into five categories: statistical information-driven, spatial statistics-driven, optimization decision-driven, mechanistic simulation-driven, and intelligent fusion-driven. The fundamental principles, typical applications, applicable conditions, advantages, and limitations of each category are systematically summarized and compared. The development of groundwater monitoring network optimization from network evaluation to optimal design and from single-method applications to integrated methodologies is further discussed. Current research still faces challenges including limited monitoring data, difficulties in multi-source information integration, insufficient uncertainty quantification, and limited interpretability of intelligent models. Future research should emphasize the integration of physical mechanisms with data-driven approaches and promote the application of digital twins, real-time dynamic optimization, and intelligent decision-making technologies to support the design and management of groundwater monitoring networks.