The ski manufacturing industry is characterized by short product life cycles and high innovation pressure in order to meet customer's expectations of progressive improvements of skis. To reduce development time a FE model was developed and validated to simulate the influence of changes of the ski construction on its bending and torsional stiffness. Moreover, the study aimed to evaluate the feasibility of a novel ski sandwich construction with a load dependent bending stiffness. Therefore, the FE model was used to define required strain-stiffening of a single construction layer material to reach a perceptible change of the overall ski's bending stiffness. Two existing skis with identical geometry but distinctive differences of their torsional and bending stiffness were modelled using ANSYS Workbench. The sandwich construction was modelled by 18 solid bodies with bonded contacts: core, sidewalls, edges, upper and lower face with 13 layers including 3 additional resin interlayers. Orthotropic linear elastic material properties were taken from data sheets of suppliers and from material tests of the manufacturer. Sweep meshing was used to create an all-hex mesh of solid elements. Static structural simulations of a 3-point bending test and a torsion test were run in accordance to existing laboratory tests to validate the model. For both skis, the FE model showed good agreement with the experimental data for ski A and B. The simulated overall bending stiffness (center spring constant) was, respectively 2.8% and 3.2% higher compared to the experiments. The elastic curves revealed the model as slightly too stiff at the afterbody and too soft at the forebody. The simulated torsional stiffness (torsional spring constant) was, respectively 2.1% and 4.2% lower than found in the experiments. In the development process, for example, the model was then used to quantify the influence of edge profile heights on the ski's overall bending stiffness (3% per 0.1 mm edge height). The application of strain-stiffening materials to realize a load depending ski stiffness turned as not feasible due to too small strains within the ski structure. A realistic representation of the different construction layers of an alpine ski is still challenging, especially due to the heterogeneity of the fibre compound layers and the resin distribution. Using bulk properties of the upper and lower face of the sandwich construction is not an alternative. To virtually test the influence of new materials and layups every single layer has to be represented. Using two skis for validation and additional resin interlayers for calibration appeared appropriate in order to achieve adequate model results.