This paper deals with a critical assessment and a reestimation of the "non-accelerating in ation rate of unemployment" (NAIRU) for Germany. There are quite a few obstacles to perceiving the NAIRU as an understandable and easy-to-use analytical instrument, suitable for economic policy: the possibility of a non-vertical Phillips curve (e.g. in times of low in ation), the occurrence of shocks and hysteresis effects, the (mis-)measurement of important variables such as in ation expectations, cointegration issues, and a time variability of the NAIRU and its confidence intervals. Despite many serious caveats a new attempt is made to estimate a NAIRU for Germany based on conventional Phillips curves as well as on new approaches such as using direct measures of in ationary expectations, the Kalman filter method, and the residual-based bootstrap procedure (in order to estimate confidence intervals). However, by any method, simple or complex, the NAIRU is very hard to determine and subject to considerable arbitrariness.