The reported clinical results following revision of failed cemented hip arthroplasties have varied. The most improtaint aspect of revision surgery, especially in case of removal of well injected femoral cement need not be daunting if certain essential principles are understanding. The prohlem is hasically that of a tuhe (cement) fixed within another tube (femur). Otherwise, any attempt to work between the two will cause the instruments to skid off the cement and penetrate the femoral cortex. Therefore, on the first step to deal with the problem and theoretical ground, we have investigated the mode of fracture in bone cement. Prior to this study, a comparision of various cement strength been published by A.J.C. Lee, University of Exetcr, UK is rviewed for our reference. As far as cement crack is concerned, the re are three modes of fracture, namely, mode I, mode II and mode III. As seen in the Fig. 5, mode I fracture is defined to be the fracture under symmetic loading, which is perpendicular to the crack surface. Mode II fracture is the fracture under anti-symmetric loading, which is parallel to the crack surface. In mode III, the loading is perpendicular to both crack surface and the plan of the paper. On the left is the situation shown when using osteotome to crush the cement Fig. 6. The prohlem here is of dynamical nature, however, this kind of prohlem has not been solved yet. Prediction of the direction of propagation is possible by assuming the static nature when the osteotome just gets inside the cement. Small element with distance and angle from the osteotome tip under stress is also,shown on the left, Here, th mode of fracture is mode I. ∂θθ is obstained from fracture mechanics as this, and the direction of crack propagation is given by solving ∂θθ = o,i.e, the direction of maximum ∂θθwhich turns out to be 0=0. When chisel is used, the mode of fracture is th combined mode, i. e. Mode I + Mode II (Fig. 7). In an angiogous approach to the previous one, aθθis obtained, thus, the direction of propagation is given by solving$\frac{\partial \theta \theta }{\partial \theta }$= 0. which results in θ=(−)α, the half of the included angle of the chisel, Note, however, that the direction of propagation given here is the initial direction of crack propagation. In practiee, the crack often gets curves, however, this phenomena is governed by other factors neglected in this presentation, such as inhomogenity, state of stress of bone, cement and stem, etc, Thus, quantitative study is required in addition to so've these phenomenon.