Bone age assessment defined as the measure of skeletal development is most often used in pediatrics and forensics to estimate the true age of a person. It is usually done by comparing the left hand X-ray of a person with the hand radiographs in the standard atlas or based on local regions of interests (ROI) that include epiphyseal regions of the phalanges (14 ROI’s).Both these assessments were labour intensive, prone to discrepancies and can only be used to estimate the age till 18. Hence there is a need to develop automated method to assess the bone age by exploiting the appropriate features. This paper attempts to identify a procedure in recognizing the respective bone that belongs to male or female with its corresponding age. The automated procedure comprises of segmentation of metacarpals using area based statistics followed by typical feature extraction. Nine features are extracted for the experimental study. A back propagation neural network is then applied to classify whether the given sample refers to male or female bone. It is observed from the simulation results that the proposed procedure is found to be less computation burden and the results are found to be comparable with the existing work reported in the literature.