Photon counting spectral x-ray CT provides powerful techniques for analyzing various materials in-situ by utilizing multiple separate energy windows. In the recent few years, preliminary results of humans and animals which were scanned with such prototype systems have been shown. Although photon counting technology allows numerous separate energy windows, practical system design should consider and optimize the advantages of additional windows relative to the complexity, image quality and system’s cost. In this work we investigated experimentally intricate material differentiation with dedicated multi energy window spectral analysis method. Particular attention was given to the simultaneous presence of several mixtures and partial volume regions of different materials. These scenarios resemble complicated analyses in potential medical and pre-clinical applications. The results show that multi energy window setting has a great advantage over dual-energy setting especially where multiple material mixtures should be differentiated one from each other.