The most widely used steel grade for rolling bearings is based on a steel composition first used almost a hundred years ago, the so-called lC-1.5Cr steel. This steel is used either in a selective surface induction hardened condition or in a through hardened heat treated condition, both yielding exceptional structural and contact fatigue properties. The Lundberg and Palmgren rolling bearing life prediction model, published in 1947, was the first analytical approach to bearing performance prediction, subsequently becoming a widely accepted basis for rolling bearing life calculations. At that time the fatigue life of rolling bearings was dominated by the classical sub-surface initiated failure mode. This mode results from the accumulation of micro-plastic strain at the depth of maximum Hertzian stress and is accelerated by the stress concentrations occurring at the micro internal defects. In common with all fatigue processes, rolling bearing failure is a statistical process: the failures of bearings with high inclusion content tested at high stress levels belong to the well-known family of “Weibull” distributions. Steady improvements in bearing steel cleanliness due, amongst other things, to the introduction of secondary metallurgy steel making techniques, have resulted in a significantly increased rolling bearing life and load carrying capacity. In recognition of this, in 1985 Ioannides and Harris introduced a new fatigue life model for rolling bearings, comprising a more widely applicable approach to the modelling of bearing life based on the relevant failure mode. Subsequently this has been extended to include effects of hardness and of micro-inclusion distributions in state-of-the-art clean bearing steel.