Many communities around the world are facing increasing flood-induced damage to bridges due to climate change and rising urbanisation. It is thus crucial to understand how different bridge types suffer from flooding and how this may affect surrounding networks. Despite the large body of literature for seismic and hurricane taxonomies, few classifications exist for bridges at flood risk. In this work, existing global bridge classifications were reviewed in order to derive a bridge–flood taxonomy. The review revealed that existing studies mainly classify bridges according to the superstructure material, whereas subclasses consider superstructure and substructure components. A taxonomy of 20 attributes for riverine roadway bridges susceptible to flood hazards is proposed in this paper. Its applicability for three bridge datasets in the UK was verified. The results showed that the considered datasets have data for 13 attributes, which can be used to derive regional bridge classes. In general, the taxonomy is functional for standardising different bridge datasets and applying/developing damage models for given bridge portfolios of flood-prone countries. Future works could apply the taxonomy to additional bridge datasets within a network for risk assessments; the proposed taxonomy could also be extended to allow integration with functionality and restoration models.