Under certain cutting conditions carbide tools can sustain a significant amount of permanent deformation and this may cause early tool failure. Tests were devised to investigate the deformation of three different grades of carbide, when machining steel (817M40, EN24) under a wide range of conditions. Each test was carried out on a continuous (60 seconds) and an incremental (5, 5, 10, 20 and 20 seconds) basis. This plan was adopted to investigate transient effects. During each test the cutting forces were measured with a dynamometer, and the boundary temperatures were measured at the tool/chip interface and the tool/shank interface. The permanent deformation took the form of bulging on the flank face and depression of the rake face. With low metal removal rates the deformation was minimal and the cutting edge was stable. High metal removal rates caused the tool to deform continuously and this gave an unstable cutting edge. It was deduced that within the tool there was a zone of material that had undergone plastic deformation. The rake and flank faces formed two external boundaries of this zone, the remaining boundary being within the tool body. For any particular set of cutting conditions, the amount of deformation for either the continuous test or the total of the incremental tests was essentially the same. A plane stress Finite Element (F.E.) model was developed to explain the effects of speed and feed in terms of temperature and stress and their variation with time. The F.E. model predicted that the values of both the transient and steady state thermal stresses were very low when compared with the mechanical stresses. The results from the cutting tests and the F.E. model suggest that the tool material continuously deformed under the applied mechanical stresses (cutting forces). Any contribution to the deformation from the transient thermal stresses was minimal and of a short duration.