In this paper, we have performed 3D finite element calculations of thin-walled tubes subjected to dynamic twisting to investigate the effect of porous microstructure on the formation of shear localization bands under simple shear conditions. For that purpose, we have incorporated into the finite element model the porous microstructures of four different additive manufactured metals – aluminium alloy AlSi10Mg, stainless steel 316L, titanium alloy Ti6Al4V and Inconel 718 – for which the void volume fraction varies from ≈0.001% to ≈2%, and the voids size between ≈6μm and ≈110μm (Marvi-Mashhadi et al., 2021). For each microstructure, we have created up to 10 realizations varying the spatial location of the voids and the distribution of voids size. The matrix material is elastic/plastic, with yielding defined by the von Mises yield criterion and associated flow rule. The yield stress evolution is considered to be dependent on strain, strain rate and temperature, with parameters corresponding to Titanium and HY-100 Steel, taken from Molinari (1997) and Batra and Kim (1990), respectively. Moreover, we have assumed the deformation process to be adiabatic. The calculations have been performed for shear strain rates ranging from 100s−1 to 10000s−1. To the authors’ knowledge, this is the first study ever that simulates dynamic torsion testing of porous materials with actual representation of voids, providing new results which bring to light the influence of porosity on dynamic shear banding under simple shearing. Namely, the numerical calculations have shown that both the location of the shear band and the critical strain leading to the shear band formation depend on the spatial and size distribution of the voids in the specimen, evidencing the influence of material defects on the localization pattern. Notably, the shear band nucleation strain decreases with both the void volume fraction in the specimen and the size of the voids, the size of the largest pore being the main microstructural feature controlling the loss of load carrying capacity of the specimen. In addition, we have carried out a parametric analysis varying the temperature and strain rate sensitivities of the material, and the loading rate. For the strain rates investigated, increasing the loading speed leads to a mild decrease of the shear strain leading to shear band formation, while the strain rate sensitivity is shown to stabilize material behavior and delay localization. Moreover, the numerical results have made apparent that for the hardening materials considered, thermal softening is essential to trigger the shear band formation, so that the porous microstructure alone does not lead to shear localization.