In dual energy x-ray absorptiometry (DXA) the photon attenuation is assumed to be similar in soft tissue overlying, adjacent to and inside the measured bone. In the calcaneal dual energy x-ray laser (DXL) technique, this assumption is not needed as attenuation by soft tissues at the local bone site is determined by combining DXA and heel thickness measurements. In the present study, 38 subjects were measured with DXL Calscan, Lunar PIXI and Lunar DPX-IQ DXA instruments and Hologic Sahara ultrasound instrument, and the performance and agreement of the instruments were analysed. Furthermore, numerical simulations on the effect of non-uniform fat-to-lean tissue ratio within soft tissue in heel were conducted. In vivo short-term precision (CV%, sCV%) of DXL Calscan (1.24%, 1.48%) was similar to that of Lunar PIXI (1.28%, 1.60%). Calcaneal areal bone mineral densities (BMD, g cm(-2)) measured using DXL Calscan and Lunar PIXI predicted equally well variations in BMD of femoral neck (r2 = 0.63 and 0.52, respectively) or lumbar spine (r2 = 0.61 and 0.64, respectively), determined with Lunar DPX-IQ. BMD values measured with DXL Calscan were, on average, 19% lower (p < 0.01) than those determined with Lunar PIXI. Interestingly, the difference in BMD values between instruments increased as a function of body mass index (BMI) (r2 = 0.17, p < 0.02) or heel thickness (r2 = 0.37, p < 0.01). Numerical simulations suggested that the spatial variation of soft tissue composition in heel can induce incontrollable inaccuracy in BMD when measured with the DXA technique. Theoretically, in contrast to DXA instruments, elimination of the effect of non-uniform soft tissue is possible with DXL Calscan.