For patients considered at risk for osteonecrosis (ON) of the femoral head, an algorithm for the efficient and appropriate use of diagnostic tests can be developed. Such an algorithm requires a clear idea of the evolution of the disease process and an understanding of the limitations of each of the modalities by which the disease process can be identified. The role of each of the diagnostic tests available for the diagnosis of ON and the results of prospective evaluations to assess their sensitivity, specificity, and predictive values are reviewed. No single diagnostic test is 100% accurate in the diagnosis of the disease. Thus, one must consider an algorithm that accounts for the variable nature of the disease presentation. The goal of such an algorithm is the accurate identification and staging of the disease. For disease not roentgenographically apparent, an approach using magnetic resonance imaging and bone scanning is recommended. For disease that is apparent roentgenographically, tomography is used to define the extent of the disease and the presence or absence of subchondral fracture. The accuracy of diagnosis is critical to understanding the role of the various treatment modalities prescribed for ON. For this purpose, it is necessary to reach a consensus as to what constitutes a successful outcome of treatment. Once diagnosis of the disease is accepted, clinical success of treatment outcomes should be reported in three ways: by a standard hip rating system, by ability to prevent roentgenographic deterioration, and by survival of the femoral head (the absence of further treatment intervention).(ABSTRACT TRUNCATED AT 250 WORDS)