Spent fuel dry cask storage facilities are proven part of many waste management concepts. The recent German strategy foresees a storage duration of a maximum of 40 years. However, the lack of a final disposal will lead to prolongations of the storage time beyond this limit. This requires detailed investigation of the long-term degradation mechanisms to ensure the safety functions to be fulfilled. Further important aspects are transportability and manageability of the fuel assemblies. This requires integrity of fuel rods and especially the exclusion of systematic fuel rod failures. A key variable to predict the integrity of the fuel rod cladding is the temperature. Almost all degeneration mechanisms are temperature dependent. A realistic prediction of the time dependent temperature field inside the cask is needed to investigate long-term degradation mechanisms properly. Overly conservative temperature estimates with the purpose to assure peak cladding temperatures to be within technical specifications might lead to non-justified conclusions.In this work we discuss the 3-dimensional cladding temperature fields in a generic Castor-like cask for different loading schemes. We generated a generic cask model similar to the GNS CASTOR V/19® for the COBRA-SFS code. In a first step, we verified the model by comparison with similar analyses performed by two other codes (ANSYS CFX and COCOSYS) and the modelling used in these analyses; all assumed homogeneous loading. We found large temperature gradients of 100K within one fuel assembly and approximately 50K variation on the fuel rod height. With these more detailed temperature fields, our work builds the basis for a more reliable prediction of the cladding behavior for a given cask during long-term interim storage.