Recently, fan beam (FB) designs have been made available by several manufacturers (Aloka, Hologic Lunar and Sophar) to measure lumbar spine bone mineral area density (BMD) in both an anteroposterior (AP) and a lateral projection. The present study was performed to evaluate some characteristics of a new dual energy X-ray absorptiometry (DXA) system for supine lateral scans in normals and to study possible advantages for patients with osteophytic calcifications (OC). The precision errors of in vitro and in vivo measurements were estimated by an anthropomorphic phantom and in healthy volunteers. To study the effect of osteoarthritic changes on AP and lateral DXA measurements, BMC (bone mineral content) and BMD were measured in age-matched women (n = 150) with and without OC. Precision errors for lateral BMD in vitro over 1 and 6 months were 0.58/0.67% (slow/fast scan modes) and 0.67/0.77% (slow/fast scan modes), respectively. The short- and mid-term reproducibility of BMD values were 2% and 3.5%, respectively, using the compare function (3.5% and 7.5%, respectively, without the compare facility). The analysis of women with and without OC (n = 150) demonstrated higher mean values for AP BMD (0.892 +/- 0.145 g/cm2) in patients with OC (n = 75) than in normals (0.836 +/- 0.135 g/cm2, n = 75, difference 6.3%). For lateral scans, BMD differed to a minor degree (3.1%) in patients with OC (0.629 +/- 0.133 g/cm2) compared with normals (0.610 +/- 0.117 g/cm2). Corresponding results were obtained in fast FB mode. Furthermore, we found significant (P < 0.0001) correlations between BMD in lateral and AP scans in patients without OC (r = 0.63) and in patients with OC (r = 0.57). Although the FB design facilitated fast AP and lateral scans, the higher precision errors of lateral scans could limit its application in longitudinal studies. The use of compare function should be recommended. However, BMD of lateral scans was less influenced by OC.