Abstract: Objective To provide a novel insight in bone tissue sampling. Methods Twenty femoral heads were extracted from sample bank of OP patients with femoral neck fracture, and they were matched with the feature of female, age gap less than 5 years and similar size of femoral head (diameter differ less than 2 mm) and were randomly assigned to two groups (A and B). Subchondral bone columns were then extracted from these femoral heads using two different methods (A for standard method while B for conventional method). The morphometric characterization of column specimens, the axis deflection angle and density parameters were compared between two groups. Results Trabeculae extracted in conventional way displayed a heterogeneous distribution as long as multiple lacerated ends. The axis deflection angle (ADA, α°) of conventional group was significantly greater than that of standard group (P=0.006). The density parameters in inferior layers were significantly higher than those in superior layers in both groups. The coefficient of uniformity of standard group was significantly better than that of conventional group in bone mineral density (BMD, P=0.000 8), bone/trabecular volumn (BT/TV, P=0.003 4), trabecular thickness (Tb.Th, P=0.000) and trabecular number (Tb.N, P=0.002 9). Conclusion The inconformity of the spatial position caused by conventional sampling method may affect the evaluation of trabecular microstructure changes, which should be considered as an important experimental bias that will influence the reliability of the results.