Vertebral bone density of two human vertebral specimens was evaluated using dual-anergy X-ray absorptiometry and single-energy quantitative computed tomography immediately before and after 4 weeks of fixation. The repeated QCT assessment of these segments produced a mean rate of change of 1.17 HU/month, while for DXA the mean rate of change was -0.002 g/month for BMC and -0.0004 g/cm(2)/month for BMD. Ten fresh and 10 fixed sheep lumbar spines underwent absorptiometry before being sectioned into units comprising a central vertebra and adjacent intervertebral discs, plus 1 cm of the flanking vertebral end-plate for embedding into bone cement. The fixed sheep spines underwent a repeat scan after 4 weeks in 10% formalin. Mechanical testing produced a significantly lower average failure load for unfixed specimens (9.3 kN) than for the fixed material (10.8 kN). Failure strain was not significantly different between groups. Linear regression showed a high correlation of BMC values before and after 4 weeks of formalin fixation, while the slopes of the regression for BMD and failure load of both fresh and fixed groups were not significantly different. While formalin fixation may result in a slight increase in compressive strength this does not appear to be associated with a systematic change in mineral density.