Bone is among the main by-products of the meat and dairy livestock. Currently, in many countries, it is valorised in pet food, disposed through incineration or exported. However, thanks to its biological and mechanical properties, it could be used in several bio-based products such as coatings, packaging, phosphorous, calcium and magnesium supply, flame retardants, etc. Successful valorisations routes of discarded bones need a comprehensive characterization of the bone tissue variability with major biological factors such as age of the animal and the anatomy of the bone itself. In this study, this variability was assessed through Terahertz time-domain spectroscopy and energy dispersive X-ray microanalysis. Bones optical properties (refractive index and absorption coefficient) and main minerals (calcium and phosphorous), were determined for adult cows’ femurs and tibias, from the same breed and breeding system and in the age range 93–120 months. Clustering analysis was then carried out, by unsupervised statistical tools (Principal Component Analysis and Agglomerative Hierarchical Clustering). Results showed that the bones considered can be grouped in two classes, represented by each anatomy; besides, the anatomy was most significant than age. The refractive index increased with age and was higher for femur; the cluster of tibias was more homogeneous than femurs. The THz radiation showed to be a useful and non-destructive tool to assess bovine bones variability and also their optical properties. This radiation could be applied to the predict animal bone components (collagen and minerals) extractability so as to valorise each cluster according to its potential.