Chemical hazardous material (Hazmat) transportation has become a very important safety issue to human society and the environment. Great attention has been drawn to reduce potential risks and incidents. In this paper, a new methodology is proposed for the dynamic routing optimization of the chemical Hazmat transportation, which includes four major stages: (i) information collection and preparation; (ii) modeling and solving individual and system routing models; (iii) reactive routing optimization under uncertainties; and (iv) trade-off study for potential shipping delays. A novel MILP model has been developed to determine the optimal shipping path with the minimal transportation risk. This model consists of two parts: the individual and system routing models, which are designed to explore the optimal shipping path for each shipping pair and all transportation tasks, respectively. When uncertainties occur, reactive routing optimization will be performed for handling the leftover transportation tasks. In particular, if some preset shipping time limits are violated due to severe uncertainties, optimal solutions subject to different allowable shipping time (AST) will be iteratively identified, so that the relation between AST and the corresponding transportation risk can be figured out. The efficacy of the developed model has been demonstrated by three case studies. Through the developed model, a reduction of 46% and 34% on system transportation risks of studied transportation tasks can be accomplished when dealing with uncertainties. The obtained relation can be used to trade off AST and the transportation risk, so that stakeholders can be advised on their decision-making for shipping path selection.