The aim of this article is to present the optimization of a proof test procedure of ceramic hip joint ball heads. The proof test rejects defective samples in the production line before being implanted into human body. Thereby on every ceramic ball head a static load is applied, which is somewhat higher than the maximum physiological load. The magnitude of the applied load should not damage the samples which are free of flaws in the high stress area. The configuration of the proof test influences the stress distribution in the ball head, which should be similar to the physiological case. To determine the stress distribution, a non-linear finite element (FE) analysis was performed and the results were validated by measurements. With an iterative approach based on FE calculations the proof test configuration was optimized in such a way that the stress distribution in the ball head is similar to the stress distribution in vivo. In this study all ball heads showed very high fatigue resistance after being proof tested and fulfilled the requirements of the FDA (Food and Drug Administration, USA) described in the Guidance Document for the Preparation of Premarket Notifications for Ceramic Ball Hip System. The probability of a fracture of an implanted ceramic ball head can be decreased by the presented optimized proof test procedure. Latter can thus improve the reliability of ceramic hip joint ball heads. The study was supported by the KTI (Commission for Technology and Innovation, Switzerland).