Deformation in the Earth's outer shell is mostly localized into narrow high-strain zones. Because they can have displacements up to several hundreds or thousands of kilometres, they can affect the entire lithosphere. The properties of high-strain zones control the kinematics and dynamics of our planet, and are therefore of key importance for an understanding of plate tectonics, stress accumulation and release (e.g. earthquakes), mountain building, etc.One of the requirements of shear zone formation in ductile rocks is localized strain softening (Hobbs et al. 1990). In this paper we review the strain softening mechanisms that were identified and proposed 25 years ago and analyse their relevance in light of recent experimental results conducted to large strains. For this purpose, some of the newer developments in experimental deformation techniques that permit high strain in torsion are summarized and recent results are reviewed. Using these results we discuss mechanisms, processes and conditions that lead to localization.Since the first deployment of GPS (Global positioning system) to monitor plate motion in the al. 2003). In contrast, the local strain rate at plate boundaries is several orders of magnitude higher and is restricted to narrow zones, even within more diffuse plate boundary zones such as central Asia. Since displacements along these high strain rate zones are of the order of hundreds of kilometres or more, they must affect the entire lithosphere. Paterson (2001) pointed out that to model the theology of the crust, one should consider it to be divided into blocks separated by faults (upper crust, where deformation is timeindependent) and shear zones (lower crust or even mantle, where the deformation is timedependent). The bulk rheology depends predominantly upon the behaviour of these shear zones.In trying to summarize and define the status quo of our knowledge about the formation and mechanical properties of shear zones it is helpful to take a look at a prescient summary published a quarter of a century ago by White et al. (1980). This and other publications we refer to were the result of a conference on shear zones in rocks held in Barcelona, Spain in 1979. We will follow the arguments of White et al. (1980) and include other publications which attempted to define the status quo at the time, for example the Penrose conference report on mylonites by Tullis et al. (1982).Although the general topic is broad, we will focus on one area, the formation of ductile shear zones due to localized relative strain softening. The questions posed at that time are discussed in the light of new results. Other important issues related to shear zones include reactivation, brittle faulting, and the role of high-strain zones as conduits for fluids, to name but a few, but these will only be addressed in so far as they affect the question of deformation mechanisms in ductile high-strain zones.