Morphometric techniques, which use conventional lateral spine radiographs to quantify vertebral body shape (morphometric radiography, MRX), have proved a useful tool in the identification and evaluation of osteoporotic vertebral deformities. Recently a new method of acquiring the images required for vertebral morphometry using dual-energy X-ray absorptiometry scanners (morphometric X-ray absorptiometry, MXA) has been developed. In this study we compare repeat analysis precision of vertebral height measurement using MXA and MRX. Twenty-four postmenopausal women were recruited (mean age 67 +/- 5.8 years): 12 normal subjects and 12 with osteoporosis and vertebral deformities. Each subject had a MXA scan and lateral thoracic and lumbar radiographs at a single appointment, which were each analyzed quantitatively in a masked fashion, using a standard 6-point method, twice by one observer and once by a second observer. Anterior (Ha), mid (Hm) and posterior (Hp) vertebral heights were measured and wedge (Ha/Hp) and mid-wedge (Hm/Hp) ratios calculated for each vertebral body. Intra- and interobserver precision were consistently poorer in MXA compared with MRX in both normal subjects and those with vertebral deformities, with MXA CV% generally at least 50% higher than corresponding values for MRX. For both MXA and MRX interobserver precision was clearly poorer than intraobserver precision, a problem associated with any morphometric technique. MXA intra- and interobserver precision were significantly poorer for subjects with vertebral deformities compared with those without, with a CV% for deformity subjects up to twice that of normal subjects. Conversely, MRX showed little or no obvious worsening of intra- or interobserver precision for deformity subjects. Comparison of MXA precision in the normal and deformed vertebrae of the deformity subjects demonstrated that the poorer precision in these subjects compared with normal subjects was the result of increased variability in point placement on the deformed vertebrae themselves. However, the precision for normal vertebrae in these subjects was also somewhat poorer than the precision in normal subjects. We conclude that MXA precision is generally poorer than that of MRX and that the presence of vertebral deformities has a more pronounced effect on MXA precision than on MRX precision.