Crack formation and development in reinforced concrete structures is a complex mechanism affected by a large number of factors. Typically, the estimation of the crack width as a function of the reinforce-ment stress state is based on simplified assumptions concerning the rebar-to-concrete interaction. In this paper, the results of concrete ties and pull-out tests instrumented with Fibre-Optic Measurements and Digital Image Correlation are used to review current code formulations. The results indicate that the assumption of average bond strength could be improved and adjusted based on visible deteriorations on the concrete surface. More refined models could be beneficial to establish a better correlation between crack width and stress in the reinforcement.