In the second half of 2011, Thailand has recently faced with the most devastating flood of her modern history. More than 3,330 national highways are damaged by this flood. Some roads are heavily destroyed while others are partially damaged and emergency relieves cannot access to the flooded areas. Highway restoration is an urgent responsibility of road authorities. This study presents the sequential highway network restoration decision model where highways are restored one by one in sequence. To determine an optimal restoration sequence, the model is formulated as a dynamic program where the primary objective is to sequentially restore roadways to minimize the travel demand loss for the disconnected network. Once the network is connected, the secondary objective is to sequentially restore roadways to minimize the network travel time where traffic assignment onto the network is based on User Equilibrium concept. A sample network is examined to investigate the solution characteristics. It is found that the proposed algorithm can provide good practical solutions to the sequential highway network recovery problems.