The 'local scaling' hypothesis, first introduced by Nieuwstadt two decades ago, describes the turbulence structure of the stable boundary layer in a very succinct way and is an integral part of numerous local closure-based numerical weather prediction models. However, the validity of this hypothesis under very stable conditions is a subject of ongoing debate. Here, we attempt to address this controversial issue by performing extensive analyses of turbulence data from several field campaigns, wind-tunnel experiments and large-eddy simulations. A wide range of stabilities, diverse field conditions and a comprehensive set of turbulence statistics make this study distinct.