The performance of MPACT available in VERA-CS 4.1RC1 and its libraries was assessed by comparison with other calculation schemes already available at PSI such as CASMO-5 and Serpent 2. A range of pincell and assembly models corresponding to a modern BWR design were considered. The accuracy in terms of k-inf, fission rates and isotopic compositions was analysed for steady state and depletion calculations. Overall no large deviations are observed between CASMO-5, Serpent 2 and MPACT for the quantities considered which gives confidence that VERA-CS could be used already at PSI for an accurate modelling of BWR problems. Nonetheless, a number of limitations are reported. The performance of MPACT tends to degrade with an increasing void content. This is most likely a library issue, related to the SPH factors tailored for PWR analysis. The burnup calculation scheme used in MPACT is less efficient than the one used in Serpent 2 or CASMO-5. Spatial trends of the fission rate differences with Serpent 2 are visible at the assembly level in all void conditions. This issue is not specific to MPACT (CASMO-5 does not perform better). Extra geometry modelling features are required for the modelling of modern BWR lattices as the current simplifications led to fission rate discrepancies in the same range as the code-to-code discrepancies. Finally, the results obtained with the latest BWR specific developments of MPACT are very encouraging and show much better agreement with the reference solution than the current VERA-CS version available at PSI.