Control rod withdrawal in BWRs induces large power steps in the adjacent fuel assemblies. This paper investigates how well a 2D/3D method, e.g., CASMO5/SIMULATE5 computes axial pin power gradients adjacent to an asymmetrical control-rod tip in a BWR. The ability to predict pin power gradients accurately is important for safety considerations whereas large powers steps induced by control rod withdrawal can cause Pellet Cladding Interaction. The computation of axial pin power gradients axially around a control rod tip is a challenging task for any nodal code. On top of that, asymmetrical control rod handles are present in some BWR designs. The lattice code CASMO requires diagonal symmetry of all control rod parts. This introduces an error in computed pin power gradients that has been evaluated by Monte Carlo calculations. The results show that CASMO5/SIMULATE5, despite the asymmetrical control rod handle, is able to predict the axial pin power gradient within 1%/cm for axial nodal sizes of 15–3.68 cm. However, a nodal size of 3.68 cm still causes underestimations of pin power gradients compared with 1 cm nodes. Furthermore, if conventional node sizes are used, ∼15 cm, pin power gradients can be underestimated by over 50% compared with 1 cm nodes. The detailed axial pin power profiles from MCNP are corroborated by measured gamma scan data on fuel rods irradiated adjacent to control rods.