Tunnel excavation represents a particularly challenging problem from the design point of view, mainly due to the uncertainties characterising the soil or rock mass behaviour and to the difficulties in the prediction of the stress state on the structural supports. This explains why tunnel designers mostly rely on observational data, adapting excavation sequences and support measures during the construction phase (at least in the case of traditionally excavated tunnels), especially for deep tunnels, where the geological and geotechnical uncertainties can be significant. However, the observational method requires a reliable monitoring system as well as considerable work to interpret and back-analyse data measured during construction. On the other hand, this interpretation and back-analysis represent an important step in gaining better understanding of the system response and allowing improvement of the design, thus avoiding excessively conservative and uneconomical solutions. In this paper, a back-analysis of stress and displacement measurements is presented for a NATM tunnel excavation in rock, carried out using a rheological model for shotcrete. It is shown, consistently with the previous literature on the topic, that the use of an advanced model improves the prediction of tunnel behaviour compared to the common practice of using quasi-elastic E-moduli.