Vehicle Routing and Scheduling (VRS) constitute an important part of logistics management. Given the fact that the worldwide cost on physical distribution is evermore increasing, the global competition and the complex nature of logistics problems, one area, which determines the efficiency of all others, is the VRS activities. The application of Decision Support Systems (DSS) to assist logistics management with an efficient VRS could be of great benefit. Although the benefits of DSS in VRS are well documented, however in practice many organisations perform these activities manually using combination of skills, intuition and expertise. A comprehensive review of literature revealed several drawbacks in the existing methods for addressing VRS. The traditional optimisation approaches have very limited applications and these require high computation time. Also, heuristic approaches are capable only to specific variation, a slight difference in the structure of the problem make the algorithm inefficient. Furthermore, metaheuristics methods require higher computation time and they are context dependent. Also, further investigations on the VRS problem formulations suggest that heuristic approaches usually address a single objective of distance minimisation. However in the real world there may be a number of conflicting objectives. In general, there is a lack of considerations for route selections, resource utilisation, unhlfilled demands, underused capacities, reliability of deliveries, fleet size, human fitness and operational cost. Also, these approaches fail to realise non-linearity within objectives and constraints defined for VRS problems. Furthermore, there are no clear distinctions between hard and soft constraints considered in these methods. Finally, the existing approaches fail to capture stochastic and dynamic nature of the logistics processes. In order to overcome the above-mentioned drawbacks, this study designed and developed a hybrid DSS to assist logistics managers with VRS tasks. The capabilities of the developed DSS have then been applied to a Liquefied Petroleum Gas (LPG) distribution company. The architecture of this DSS is composed of Genetic Algorithm (GA) optimisation tool and a simulation model. The GA module aims to provide a pool of near optimum transportation schedules. The simulation module is used to further evaluate the generated schedules. The feed back from the simulation module is used to update the GA for reoptimisation. Some unique features of this DSS are such as: development of a multi modal genetic algorithm to address VRS problems; considering supply chain performance measures as part of VRS problem formulation; allowing consideration of different objectives, soft or hard constraints concerning the supply chain, considering linearlnonlinear relationships within objectives and constraints defined and finally, considering stochastic and dynamic behaviours of the supply chain system. The GA and simulation tool integration provides unique benefits that have not been in the literature such as consideration of practical requirements, uncertainties, dynamic and stochastic behaviours, considering several criteria and producing different alternative solutions. Also, this integration allows the GA model to filter out solutions that are less competitive and therefore reducing the simulation time evaluation, which is computationally expensive. Furthermore, the human interaction with the system assists in generating higher quality of solutions. Finally, the clear benefit of this DSS is the fact that it greatly influences the applicability of the GA generated schedules and provides better confidence in implementation of these solutions