Aseptic necrosis of the femoral head is an often-disabling disease entity which can result from any of several forms of circulatory compromise to tenuously perfused weight-bearing bone. Preservation of the natural femoral head is an important goal, particularly in the younger adult patients for whom prosthesis loosening would be a strong likelihood. Presently, it is difficult to predict which infarcted femoral heads will suffer early or late collapse, and the relationship between involvement region and severity of symptoms is not well understood. We hypothesize, based on preliminary studies and data in the literature, that femoral heads with osteonecrosis of varying configurations not only have varying potential for collapse, but that collapse can be reliably predicted by appropriate mechanical models. Osteonecrosis generally involves structural comprise of affected bony regions due to the fact that the complex natural repair processes which accompany infarct revascularization tend to result in net resorption of necrotic trabeculae, prior to eventual trabecular reconstitution with newly deposited viable bone. We propose to use finite element analysis to study several aspects of the structural behavior of the weakened necrotic femoral head. A fully anatomic three-dimensional linearly elastic model is proposed (Specific Aim I) to estimate the stress aberrations associated with five general configurations of infarct development, and to estimate the mechanical consequences of core-drilling, bone grafting, and varus/derotation osteotomy for each of these infarct configurations. We propose also to exercise to three-dimensional elastic model for individual infarct configurations (Specific Aim II) detected in five specific patients using magnetic resonance imaging. To more closely explore the mechanisms of actual structural failure in the necrotic femoral head (Specific Aim III), we propose to build upon preliminary work a linear elastic fracture mechanics finite element algorithm that has proven capable of predicting crack propagation patterns similar to those seen clinically. Although the biomechanical aspects of osteonecrosis are presently not well understood, most current therapeutic measures are essentially mechanical in nature. The significance of the proposed work lies in helping better delineate the indications for surgical intervention, and in helping to more selectively identify which procedure is likely to be of greatest benefit for specific head involvement patterns.