Codes of practice can be overly conservative, particularly for the shear resistance of reinforced concrete beams with shear reinforcement when large loads act close to supports. This thesis addresses the topic by proposing a refined design approach based on the stress field method for this type of members, with the aim to provide accurate values for the design or verification of both slender and squat members. It then goes on by presenting a model to better understand the behaviour of steel and concrete in presence of dowel action in reinforcing bars. Consistent models based on suitable stress fields, as for example the Variable-Angle Truss models, can be used for the shear design of slender members with shear reinforcement, as proposed by Eurocode 2 (EN:1992-1-1:2004) and fib Model Code 2010. However, this type of approach usually neglects the contribution of direct struts for loads applied close to the supports, and thus underestimates the resistance in these cases. To account for this phenomenon, the shear resistance of slender members in design codes has typically been adjusted by empirical corrections. This thesis shows the advantages of designing these cases based on tailored stress fields, which allow a smooth transition between slender and squat members and yield more accurate predictions than empirical corrections. On that basis, simple design formulae are developed to serve as a basis for a revision for the next generation of design codes (Model Code 2020 and 2nd generation of Eurocode 2). Reinforcing bars are commonly designed to carry axial forces, neglecting their ability to resist transverse forces by dowel action, which can occur at crack interfaces, connections between various concrete elements or between two concrete parts cast at different times. On the negative side, dowel action can affect the fatigue resistance of reinforcing bars subjected to cyclic loading, inducing stress concentrations near interfaces with relative displacements transverse to the bar. This thesis contributes to a better understanding of dowel action by two test series. The first series focuses on the dowel response due to monotonic or low stress-level cyclic actions, with optical fibre and digital image correlation measurements. The results show the influence of the bar diameter, the imposed crack kinematics and the angle between the bar and the crack. The second test series investigates the behaviour of concrete underneath the bar due to a point load. The results show a strong dependency on position of the load along the bar. As for the stress prediction in reinforcing bars due to dowel action, this thesis presents a new formulation for the bearing stiffness of concrete under the bar to be introduced in Winkler's model as a function of the transverse displacement. The formulation is calibrated based on mechanical considerations and optical fibre measurements. The proposed bearing stiffness leads to good predictions of both the dowel force-transverse displacement response and the peak stress in the reinforcing bar for both monotonic and cyclic tests.