The diagnosis and classification of primary bone tumors remains as much a challenge today as it has for the last 80 plus years. Although pathology is invariably equated with the image of a diagnostic microscope, the vast majority of diagnoses are made grossly with the unaided eye, as are the tissue specimens selected for microscopic "confirmation." Radiologic studies, particularly plain radiographs, remain the gold standard in gross pathologic diagnosis of the skeleton. Today, confirmation and final classification continue as the pathologist's domain, but perhaps not for long, considering the evolving ancillary imaging techniques and progressive sophistication of magnetic resonance (MR) imaging. The bone tumor cases collected and compiled by Ernest Codman, M.D. during the second through fourth decades of this century formed the basis of the first tumor registry. The Codman Bone Sarcoma Registry demonstrated among other things the importance of radiographic/pathologic correlation, underscoring the reliability of a bone tumor's location, margin (host bone/tumor interface), periosteal reaction, and matrix patterns as an accurate guide to classification and likely future biologic behavior. "A General Theory of Bone Tumors," written by Lent C. Johnson nearly 50 years ago and published in the Bulletin of The New York Academy of Medicine (February 1953, second series, vol. 29, no. 2, pp. 164-171), provided a conceptual cellular approach to the understanding bone tumor dynamics reinforcing radiologic/pathologic correlation as a reliable diagnostic tool. At the time of Dr. Lent C. Johnson's death (1910-1998), he was literally working on an updated version of his original article, the latter of which is being reprinted as the core of this illustrated revision. Our continued experience with bone tumors over the past five decades has only served to validate, on a daily basis, the fundamental principles outlined in Johnson's original article. In like fashion, it is important to keep in mind that terminology and nomenclature has also evolved since 1953, despite a continued inability to achieve complete consensus.