Groundwater monitoring network design involves the determination of sampling locations and sampling frequencies to characterize the contaminant plume in an aquifer over space and time. This paper provides a retrospective overview of the considerable advances in the field of groundwater monitoring network design over the last two decades. The methods commonly applied to sampling network design for contaminant plume monitoring can be divided into three categories: statistical methods, simulation modeling, and simulation-optimization modeling. Selection of appropriate methods for monitoring network design is ultimately dependent on the objectives of the site-specific monitoring networks and the amount and type of available data. Furthermore, it is pointed out that the most important factor affecting the final outcome of monitoring network design is the uncertainty in the aquifer and contaminant properties both spatially and temporally. The future research should focus on exploring the application of groundwater network design theories to real-world monitoring sites and developing efficient and robust simulation-optimization software for contaminant plume monitoring.