Currently, there is no road deterioration (RD) model that can address pavement performance with flooding.Therefore, the aim of this work was to develop an RD model that considers flooding and pre-and post-flood strategies. The case study is road group(s) in Queensland, Australia, based on the 34 000 km road database of its main roads authority, as Queensland was severely affected by floods in 2011. This paper provides a brief literature review and proposes an approach to develop an RD model based on a non-homogeneous Markov transition probability matrix. The new RD models are developed for a number of road groups based on roughness and rutting data collected since 2000. Detailed results on the derived roughness-based and rutting-based RD models with flooding for two selected road groups are highlighted. The new RD models show significant pavement deterioration at different probabilities of flooding events, especially at higher probabilities. It is believed that these results will help road authorities in managing flood-damaged roads with appropriate treatments and optimum budget.