Paperboard is an important material for packaging food and beverages, with hundreds of billions of packages produced every year. Since quality of the packaging material is key, it is essential to understand the influence of fibre properties and fibre orientation on the mechanical response of paperboard. This is indeed challenging because of the natural variability in the mechanical properties of paperboard fibres. Material models that take microstructure into account in the modelling framework can be used to design the paperboard structure and predict the performance of the packages. Virtual modelling is a powerful alternative to experiments, for gaining insight into the mechanical behaviour of paperboard and predict its response in complex load cases. To this end, continuum based paperboard models have been used to predict the macroscopic mechanical response. Macroscopic models have successfully captured creasing and folding in industrial packaging applications [1]. Yet, a micromechanically motivated model is needed for a deeper insight into the micromechanical response, and enables optimization of the material designs. In this contribution, a multiscale framework is implemented, where the macroscopic response is determined entirely by a microscale statistical volume element (SVE). The SVE, which consists of fibres and void, is generated from X-ray computer tomography images of paperboard. The image intensity gradient is used to identify the structural tensors, which define the fibre orientations. Fibres are modelled as a transversely isotropic elasto-plastic material. To alleviate the inherent computational effort of multiscale analyses, an eigenstrain-based reduced order model [2] is implemented. In this reduction, the SVE is divided into partitions, where partition-wise uniform plastic strain fields are assumed. This reduces the computational effort by orders of magnitude, yet enables accurately capturing simple loading cases, such as uniaxial loading. This contribution provides a very promising framework upon which a more complex model will be developed. [1] Robertsson K., Jacobsson E., Wallin M., Borgqvist E., Ristinmaa M., Tryding J. A continuum damage model for creasing and folding of paperboard. Packaging Technology and Science, Vol. 36, 2023. [2] Fish J., Filonova V., Yuan Z. Hybrid impotent-incompatible eigenstrain based homogenization. International Journal for Numerical Methods in Engineering, Vol. 95, 2013.