Performance prediction for tunnel boring machines (TBM) is a crucial topic for the construction industry as it is one of the key factors to determine construction time. For such performance prediction models, the Uniaxial Compressive Strength (UCS) and the Brazilian Tensile Strength (BTS) of rock are used as key input parameters today. This thesis examines an alternative test, the Punch Penetration Test (PPT) for its possible applicability in performance prediction. The PPT is studied from a lot of different aspects, from the correct sample preparation to test execution to the analysis of characteristic points to the possibility for improving existing performance prediction models to failure analysis and damage assessment. The results of the performed PPT were compared to linear cutting machine (LCM), UCS and BTS tests. For the improvement and integration of the outcome into a prediction model the Colorado School of Mines (CSM) model was used as a base. The research is based on 6 eastern alpine lithologies: Augen Gneiss, Calcerous Mica Schist, Granite Gneiss, Brixen Granite, Imberg Sandstone and Schist Gneiss. These lithologies cover a wide range of different rock types, from low to high strength and from brittle to ductile. Especially the circumstance, that the PPT not only considers strength in the testing procedure, but also brittleness, makes it a powerful tool for TBM performance prediction. Additionally an extensive failure assessment program was carried out and proved the PPT to be more applicable for rock cutting problems than UCS or BTS tests because of the failure mode during a PPT which is very close to the process of rock breakage in tunneling and mining. Furthermore specimens were equipped with strain gages to investigate how stress spreads out inside the rock core. The results of this study illuminate the possibilities that the PPT could offer for upcoming tunneling projects if it was used consequently, but also the disadvantages that sample preparation, testing equipment and procedure involve in its current version.