Flood hazards can affect road networks by closing roads, extending travel distances and lengthening travel times. This paper innovatively integrates flood hazard, road network topology and vehicle vulnerability via a severity factor to assess the accessibility of exposed links and the performance of the whole network. The overall network functionality loss under different flood return periods is assessed by evaluating the functionality of each network link for cars and SUVs. The most vulnerable links are identified and used to assess the performance of the entire network using topology-based metrics such as average shortest paths and isochrones. The case study of Bristol (UK) is investigated. It is demonstrated that network status is a function of vehicle typology, with SUVs exhibiting naturally stronger resistance to flood than cars. This finding can support preparedness strategies of road networks in the face of future flood events.