In the field of Disaster Risk Reduction (DRR), there is a proliferation of research into different ways to measure, represent, and ultimately quantify differential social vulnerability to natural hazards. This study shows how the same census data but three alternative multivariate methodologies can lead to radical differences in our assessment of the most vulnerable population groups to flood risk. A case study of the 2013 floods in the Parrett catchment, Somerset (UK), provides context.