Fibre-reinforced polymer (FRP) composites have been employed in the last few decades largely for the strengthening and seismic retrofitting of existing reinforced concrete (RC) structures. Several studies are available in the literature and different analytical models have been proposed for evaluating the FRP contribution in strengthened RC elements. This paper analyses the accuracy of analytical models widely used for evaluating the flexural and shear contributions provided by the FRP. Some of those models are included in design guidelines. In particular, the analytical models for evaluating the FRP strain at intermediate crack-induced debonding failure are analysed. The accuracy of each formulation is assessed comparing the analytical provisions with the experimental results collected from two databases, one for bending and one for shear. The results obtained show that most of the analytical flexural models achieve a good level of accuracy and only a few models provide inadequate results. A new formulation proposed for evaluating the FRP shear contribution is shown to be generally conservative, which comes at the expense of accuracy.