In general, fiber reinforced polymer materials are being used in case of challenging mechanical requirements. Important parameters determining performance of such polymer composites are properties of the components as well as the quality of the fiber-polymer interface. For example, porosity or voids particularly at the interface will influence the overall performance of the composite negatively. Therefore, it is of importance to know the actual porosity value to e.g. calculate and predict the mechanical properties. In principle, a CT scan can provide the means to determine the amount, size, and shape of the voids in an X-ray transparent material. The difficulty, which arises in fully organic based i.e. carbon fiber reinforced polymer composites, is the small difference between absorption of X-rays in the carbon fibers versus the surrounding polymer matrix. At certain conditions, when single fibers need to be resolved and voids surrounding the fibers may exist, CT reconstruction artefacts introduced by standard algorithms are visible as "virtual" voids in large quantities severely limiting the correct determination of the actual porosity. On the other hand, in glass fiber reinforced polymer materials, the large density difference between fibers and matrix causes artefacts to appear, too. Specific experimental settings necessary to deal with the high absorption of the glass fibers (beam hardening enforces the use of X-ray fillers) cause very low absorption differences between air and the polymer matrix and thus hardly provide information on the existence of voids. Iterative reconstruction techniques, however, are more robust against the encountered CT-reconstruction artifacts and may enable accurate determination of important sample features in polymer composite materials, and especially allow for accurate determination of voids or porosity. This work explores the use of two iterative versus standard reconstruction algorithms (CGLS and SIRT versus FDK) and validates the results with cross-sectional Optical Microscopy. The paper shows that iterative reconstruction algorithms outperform the standard reconstruction algorithm with respect to mitigating the influence of the CT reconstruction artefacts.