No validated, generally accepted dataset on the mechanical properties of braintissue exists, not even for small strains. Most of the experimental and methodological issueshave previously been addressed for linear shear loading. The objective of this work was toobtain a consistent data set for the mechanical response of brain tissue to either compressionor shear. Results for these two deformation modes were obtained from the same samples toreduce the effect of inter-sample variation. Since compression tests are not very common, theinfluence of several experimental conditions for the compression measurements was analysedin detail. Results with and without initial contact of the sample with the loading plate werecompared. The influence of a fluid layer surrounding the sample and the effect of frictionwere examined and were found to play an important role during compression measurements.To validate the non-linear viscoelastic constitutive model of brain tissue that was developedin Hrapko et al. and has shown to provide a good prediction of the shear response,the model has been implemented in the explicit Finite Element code MADYMO.The model predictions were compared to compression relaxation results up to 15% strain ofporcine brain tissue samples. Model simulations with boundary conditions varying withinthe physical ranges of friction, initial contact and compression rate are used to interpret thecompression results.