The vario-scale concept has proved an alternative for the well known discrete multi-scale solution. However the vario-scale concept is still ongoing research and requires further development e.g. on how to generate better content for vario-scale maps? This research does focus on how to incorporate hydrography networks in the vario-scale solution and how to continuously generalise hydrography networks throughout all scales. The pre-processing of topographic data is adjusted, information on hydrography features that are below another feature is stored and hydrography line features are added. The planar partition created in the pre-processing is used for the creation of the initial topological Generalised Area Partition (tGAP). Hydrography graphs are created by iterating over all hydrography features in the data structure and storing the hydrography features as objects and the connection between two neighbouring hydrography features as links. Based on the links between objects the connected components, in this case the hydrography networks are created. There can be multiple hydrography networks in one data set, which are used for the decisions made in the generalisation process. The proposed improvements and decisions for hydrography features do result in meaningful hydrography networks throughout all scales and are an improvement to the vario-scale approach without these improvements and decisions.