This research focuses on the structural behaviour of an elastic supported reinforced concrete slab. Over the years, there is a tendency to realise industrial floors with large joint-spacings. The downside of a slab on grade with a large joint-spacing is that the risk of (uncontrolled) cracking due to a shrinkage load is higher, which often results in durability issues. In practice, the slabs on grade are generally designed and constructed based on experience by specialised floor companies and often with less than the theoretical minimum required reinforcement. This research aims to close the gap between theory and practice and to gain insight into the structural behaviour. Moreover, reducing the number of damages in slabs on grade. In the first part of this thesis, the stress development and crack risk in a restrained concrete slab on grade exposed to shrinkage is analysed. The restrained shrinkage stress in the concrete member is mainly influenced by the time-dependent properties of concrete and the degree of restraint of the slab. The most relevant degree of restraint for a jointless slab is by soil friction and depends on the slab length, soil properties and the vertical load like self-weight. From the sensitivity analysis in this research, it is shown that the middle section is fully restrained for slab lengths or joint spacings of 40 to 80 meters long for practical ranged values. The relaxation effect in concrete, based on the creep behaviour, reduces the restrained shrinkage stress over time. Stress reductions of 40% to 70% were found for practical values in the analysis after one year of shrinkage loading. The crack risk in indoor slabs on grade is the highest between two and twelve months after casting according to the findings in practice and this thesis. The early-age crack risk is relatively low for an indoor slab on grade compared to a concrete pavement because of the significant less influences of the environment. The crack risk can be reduced by composing a concrete mixture with a low relaxation coefficient, low concrete shrinkage, low Young's modulus, a high concrete cracking stress capacity and by reducing the degree of restraint of the slab. Secondly, the cracking behaviour of a fully restrained concrete tensile member with only reinforcement near the surface is analysed with analytical models and a finite element model. For modelling cracking in a restrained member initiated by an imposed deformation in the finite element model, the random field application is required. The random field application generates the spatial variation of the tensile strength of the concrete material. The random field simulates weak zones in the concrete member where crack localisation will be initiated. The downside of the random field application is that the crack pattern in the model is predetermined. The cracking behaviour is analysed for a shrinkage load, a point load and for both loads combined. In all the analyses, it is found that cracking in a slab on grade always results in a not fully developed crack pattern. A not fully developed crack pattern is characterised by the forming of additional cracks instead of widening of the existing cracks when the load increases. When cracking due to a shrinkage load in fully restrained members occurs, the imposed load decreases, which reduces the tensile force in the uncracked zone. Force reductions up to 73% were calculated after cracking in a five-meter long restrained slab. For an eccentrically top reinforced member, the first crack occurs at the bottom fibre and is uncontrolled. The tensile force in the reinforcement, in combination with the eccentrical position, induces a negative bending moment in the uncracked zone of the slab on grade. In combination with the restrained shrinkage load, this results in an axial tensile force and a negative bending moment in the uncracked zone after cracking. Over time, the shrinkage load increases and a flexural crack can be initiated at the slab surface. A point load induces a positive bending moment at the position of the load and a smaller negative bending moment next to the point load. After cracking due to the point load, an axial compressive force and an increased negative bending moment are present in the uncracked zones next to the point load. The sensitivity analysis on the behaviour by the point load showed that the principal stresses in the slab could be reduced by increasing the subgrade modulus, the slab height or by lowering the concrete stiffness of the slab. The load combination analysis showed that an axial tensile force and a negative bending moment are present in the uncracked zones. It is found that a higher point load only increased the slab deflection and the number of flexural surface cracks and not the crack width. However, a higher shrinkage strain load did slightly increased the maximum crack width but mainly increased the number of cracks. Therefore, it is concluded that the slab on grade remains in the cracking phase for the combined loading situation. Both in this research and the literature, it is stated that crack width calculations of a restrained concrete slab imposed by a shrinkage load results only in indicative values. The analysis of the cracking behaviour showed that cracking at the surface is induced by an axial tensile force in combination with a negative bending moment. Therefore, estimation of the crack width can best be performed with the CUR-65 model based on an enhanced cracking bending moment of the slab on grade. Because accurate crack width calculations are not possible, recommendations are made on how to reduce the crack widths in a jointless slab on grade. A slender slab with a low concrete tensile strength is advised to reduce the internal cracking force and bending moment. Also, it is found that a lower shrinkage load leads to a smaller crack width. Therefore, a shrinkage reducing concrete mixture is advised. Moreover, high bonding properties between the reinforcement and concrete reduce the crack width, which can be realised with reinforcement with a small diameter and mesh-size. Finally, reinforcement is the most effective in controlling surface cracks when placed in the top fibre of the slab. It is concluded that for a slab on grade, replacing the conventional reinforcement with steel fibre reinforcement does not reduce the required amount of reinforcement for controlling surface cracks.