Bone mineral density accounts for 70% to 80% of the mechanical resistance of bone but is unrelated to bone tissue structure. The vertebral fracture risk increases with advancing age irrespective of whether or not bone mineral density decreases, suggesting that changes in bone microarchitecture contribute significantly to the development of osteoporosis. In contrast to bone mass, bone architecture is difficult to evaluate. Among the various methods developed to investigate bone structure, biomechanical studies are of limited value since they are done on cadaver bones. Measurement of microarchitectural parameters (e.g., mean trabecular thickness, density and separation) in bone specimens obtained by needle biopsy is the gold-standard technique. Parameters reflecting trabecular interconnections (e.g., total number of nodes and free ends) can also be measured on needle biopsy specimens. New techniques of as yet unproven validity include star volume and trabecular bone pattern factor measurement. Noninvasive techniques capable of supplying qualitative information about bone tissue are also under study. Ultrasonography can theoretically provide data on bone microarchitecture but has not yet been proven useful in clinical practice. Statistical, structural, or fractal analysis techniques can be used to evaluate bone texture on digitized roentgenograms, computed tomography sections, or magnetic resonance imaging displays; although this approach holds great promise, it is still under evaluation and has not yet been compared with histomorphometry. Lastly, the apparent relaxation time of bone marrow determined using magnetic resonance imaging may also provide information on bone structure.