Although the RTM process is being widely used, there is a critical lack of understanding related to microstructure-proeess-property interactions. Due to this, futf advantage is not being taken of the opportunities offered by RTM, including local tailoring of reinforcement, parts integration, fabrication with tight tolerances, and designing materials for damage containment. The choice of fabric architecture and fiber volume fraction, for example, have a significant effect on the design of the tool with regard to the location and design of both the gating and venting arrangements, among other factors. Preforms have to be designed simultaneously from both the structural and resin infusion aspects. In this paper, we investigate the effect of preform architecture on flow in resin transfer molding. Previous investigations aimed at permeability determination (Adams et al., 1986) have focused on only one type of fabric architecture and used ideal fluids. Although the use of ideal fluids (like corn syrup) in flow experiments gives trends as to flow front growth and behavior, the experiments do not give the investigator critical information in that factors such as fiber wet-out are neglected. In this experiment, a vinyl-ester resin was injected into an acrylic topped tool at varying flow rates, and the effects of fiber architecture (continuous strand mat, plain weave, knits etc.), fiber volume fraction, and flow rates were investigated. Based on these experiments, design guidelines are presented and solutions suggested for achieving good wet-out and uniform flow through actual structures. The close interaction between design for mechanical performance and that for good flow is emphasized.