Background: Radiographic absorptiometry (RA) is one of the earliest methods of bone densitometry and has been used to measure the phalanges and metacarpals where soft tissue attenuation is minimal. The aim of this study was to determine whether the technique can be adapted to correct for soft tissue attenuation and measure areal bone mineral density (aBMD) in the forearm. Methods: A total of 51 patients referred for a clinical spine and hip dual-energy X-ray absorptiometry (DXA) examination and 8 young and middle-aged volunteers were recruited to this study. The first 29 participants (20 women, 9 men, aged 61 +/- 14 years) served as the training cohort, and the remaining 30 (20 women, 10 men, aged 55 +/- 16 years) comprised the validation cohort. All participants underwent a DXA scan of their non-dominant forearm, and a digital X-ray image of the same arm was acquired with a step phantom. Identical regions of interest (ROIs) in the radius and ulna at the one-third radius site were measured on the X-ray and DXA images, and a soft tissue ROI was measured on X-ray images between the radius and ulna. The X-ray measurements in the training cohort were expressed as equivalent step phantom thickness (Eq. SPT) and used to estimate forearm aBMD using a linear equation calibrated against the DXA scans. Estimates of forearm aBMD made from the digital X-ray images acquired in the validation cohort were compared with the results of the DXA scans. Results: Digital X-ray estimates of radius and ulna aBMD at the one-third radius site in the validation cohort showed a good correlation with GE-Lunar iDXA scanner measurements (r=0.795; P<0.001). The Bland-Altman plot had a mean bias of -0.002 g/cm(2) and 95% limits of agreement of -0.185 to +0.181 g/cm(2). Conclusions: Digital X-ray estimates of proximal forearm aBMD corrected for soft tissue attenuation correlated with DXA measurements with correlation coefficients comparable to those seen for other peripheral bone densitometry technologies.