Conguring a FlexRay network is very complex given the number of parameters involved and their interactions, and also given the number of, sometimes contradictory, design objectives. This paper aims to come up with methodological guidelines for configuring FlexRay. This starts with a precise identication and prioritization of the car manufacturer's objectives and requirements. For instance, it might be needed to re-use legacy software developed for CAN, to enable carry-over of ECUs, to keep the data rate as low as 2.5 or 5Mb/s for using low-cost hardware and achieve better resilience to EMI, etc. Then, sub-problems are identied, as well as the factors, or constraints, influencing the design choices: network configuration, communication cycle, static and dynamic segment conguration, frame packing, allocation of the slots to the ECUs, etc. Among these design choices, we identify the ones that should be chosen by the system designer on the basis of the intended use of FlexRay, and thus are outside the scope of optimization, and the ones that will be best decided by algorithms. In the latter case, we explain how each of the issues can be tackled in the light of the existing results.