Water quality monitoring (WQM) is crucial for managing and protecting riverine ecosystems. Current WQM network design practices often rely on unsubstantiated criteria rather than accountable algorithms. Water managers face difficulties to relate the impact of local boundary conditions on the choice of appropriate WQM network design methods. After reviewing the commonly used design methods and their resulting monitoring setups, it was evident that multivariate statistical analysis is the most frequently used method for designing WQM networks in rivers. The majority of studies reported in the literature were conducted on very large rivers and originated from high- to middle-income countries. Most commonly monitored water quality parameters cover the general physicochemical characteristics and organic pollutants, without considering the ecological quality of the river. In most studies, decision on sampling frequencies depended on expert’s judgements. Data availability and expertise seem to affect the selection of design methods rather than river size and the extent of the monitoring networks. Findings from this study support that future research should simultaneously consider all relevant aspects at watershed scale and focus more on biological indicators. In addition, comparative studies with several design methods could also help identify better selection principles.