The objective of this study was to elucidate the structure of cancellous bone and its age-related changes at different skeletal sites. Therefore, the lumbar spine, iliac crest, femur, and calcaneus of 12 age- and sex-matched skeletal healthy autopsy cases (6 females, 6 males, aged 28-84 years, mean 54 years) were removed. The following analysis includes an evaluation of the trabecular bone volume (BV/TV, %) and the trabecular interconnection (TBPf, mm-1) as well as a qualitative investigation of the structure of trabecular bone. BV/TV shows the following mean values: lumbar spine, 8.3% (+/- 0.8%); iliac crest, 11.5% (+/- 1.6%); intertrochanteric, 10.2% (+/- 1.2%); femoral neck, 15.8% (+/- 1.6%); and calcaneus, 15.4% (+/- 2.0%). There are significant differences between the BV/TV of the femoral neck and that of the lumbar spine as well as between that of the calcaneus and the lumbar spine (p < 0.01). However, a positive correlation can be seen between the bone mass of the spine and that of all other investigated sites (r = 0.67 to r = 0.80; pr < 0.1). The trabecular interconnection of the lumbar spine (2.7 mm-1, SEM +/- 0.2 mm-1) and the femoral neck (0.3 mm-1, SEM +/- 0.3 mm-1) differs significantly. Only these two sites show a significant positive correlation of TBPf (r = 0.60; pr < 0.1). Age-dependent alteration of the spine and the femoral neck in bone mass and bone structure is nearly the same. The trabecular microarchitecture of the iliac crest varies systematically. A region 4-10 cm behind and 1-3 cm below the anterior superior iliac spine turns out to be the most suitable biopsy site because of its closest relation to the lumbar bone mass. However, drawing information about the trabecular interconnection within the lumbar spine by measurement of the iliac crest at any site seems to be impossible. The horizontal specimens reveal a vertical running tubular spongiosa pattern that is arranged in concentric rings starting from the dorsal shell like a honeycomb. The comparison of TBPf in horizontal and vertical planes and its age-related changes indicates the age-related bone loss to be predominantly a loss of horizontal trabeculae. Thus, the presented data provide further information about the skeletal distribution and heterogeneity of the trabecular microarchitecture.