Reliable, predictive and patient-specific biomechanical analyses would be extremely useful for diagnosis and optimization of therapy, in the field of diseases occurring in human organs and tissues. Nevertheless, their achievement is a difficult challenge. Several biological constructs undergo large displacement and frequently moderate strains. Moreover, they are characterized by a complex non-linear mechanical behavior strongly depending on tissue histological features and biological processes. Thus, there is a great need for the development of accurate constitutive models accounting for coupled mechanisms, usually occurring at very different length scales, and involving different physics. This goal can be achieved through refined approaches able to model mechanobiological effects from the molecular nanoscale level up to the macroscale tissue, and accounting for dominant biomechanical interactions. Good predictive theories have to be founded on a few measurable histological parameters and should permit one to examine by simulation different physiological and physio-pathological states.