The water footprint has been introduced as a potential sustainability indicator for human-induced water consumption, and has frequently been studied at local, national and international scales during the last decade. While water footprints are sometimes understood as a measure that includes environmental impact assessment, the water footprint as used in this paper refers to volumes of water consumed, without including weighting procedures to allow for the assessment of impacts. Two types of approaches have been applied to calculate the water footprint in the literature: bottom-up and top-down approaches. This study compares and discusses advantages and limitations of the water footprint of nations based on two input-output top-down approaches (Water Embodied in Bilateral Trade (WEBT) and Multi-regional Input-Output Analysis (MRIO)) and of the existing national water footprint accounts from the literature based on the bottom-up approach. The differences in the bottom-up and WEBT approaches are caused by inter-sectoral cut-off, because bottom-up approaches do not consider the entire industrial supply chains, while the WEBT method covers the water footprint by tracing the whole domestic supply chain of each country. The differences in the WEBT and MRIO approaches are due to an inter-regional cut-off effect, as the WEBT approach only traces domestic supply chains whereas the MRIO approach traces entire global supply chains. We found that both bottom-up and top-down approaches are heavily dependent on the quality of existing datasets, and differ substantially. The total water footprints of nations based on different approaches vary by up to 48%, and this variation is even larger at the sector level.