Somatometric techniques are important tools in forensic investigation as well as in studying biological variation. The present study investigated the long bones of the upper extremity to ascertain which one will be more appropriate to generate regression equations for prediction of BH. Four long bone lengths (HL UL, RL and CL) were evaluated subcutaneously. BH was taken using Anthropometer, while randomly HL, UL, RL and CL were measured on the right side of the body from a pull of 211 subjects using measuring tape. The findings of the present study indicate that the correlation 'r' between BH and the right upper limb measurements were significant for HL, UL, RL, in the overall population; the highest correlation between the dependent variable BH and the independent variable (UL) in overall population was 0.82 (P< 0.0001). In the females the correlation was significant for HL UL, RL, and CL; the highest correlation was found in UL, r=0.71 (P< 0.0001) the least was indicated by CL, r =-0.25. In the males the study revealed positive significant correlation for HL UL, and RL, The highest correlation was seen in UL, r=0.71 (P< 0.0001) while the least was 0.26 (P=0.02) as seen in HL. This present study shows that the UL is a veritable dimension in the estimation of BH for the Igbos because the Pearson's correlation coefficient 'r' obtained was the heights amongst that obtained for the other segments HL, RL and CL. Also the coefficient of determinant 'R' for testing model adequacy was highest for UL which had a lower SEE.