Custom-made stems are a particular class of hip prostheses manufactured in a single sample for a specific patient. Experimental fatigue testing according to ISO standards cannot be performed for pre-clinical validation of these devices. However, the implant manufacturers need to assess the endurance properties of custom-made stems. This study investigates a theoretical protocol to predict the maximum stress induced in the stem by the ISO experimental test set-up. Stress was predicted using beam theory and finite element analysis (FEA). Strain measurements were used to assess the accuracy of the theoretical calculation. Fatigue testing was performed to verify the theoretical prediction about the fatigue stem performance. The results showed that FEA is more accurate than beam theory. Beam theory calculation is able to predict the static stresses induced by the ISO 7206/4 loading set-up with a difference always lower than 20% with respect to a prediction of a simplified FEA. Hence beam theory can be used to estimate the maximum stress. FEA becomes useful for a stem stressed at a limit condition. In both cases precise data about the endurance properties of the material are required for corrected predictions. However, endurance verification should be performed by introducing a safety factor to account for the material and manufacturing variations.