In contrast to conventional manufacturing methods for sandwich structures, in this paper a foamable polyurethane system is used to impregnate textile layers and to create the sandwich foam core in one single process step. Given this simultaneous process, a high adaptability of the resulting mechanical properties can be achieved by varying the textile reinforcement of the sandwich top layers, the foam core density and the sandwich thickness, respectively. Within this paper, a design of experiment approach is implemented by using statistical analysis of variance (ANOVA) to investigate the influence of selected geometrical and material parameters on the resulting bending properties of these novel sandwich structures. A partial factorial design helps to identify the sensitivity of individual parameters and their interaction, as well as the correlation between sandwich morphology and bending performance. Based on the gained results, statistical models will help to identify parameters with statistical significance. Therefore, optimized sandwich structures according to stress-related requirements and geometrical constraints can be designed.