Since the durability depends, among other things, on the quality of the concrete and presence of cracks, it is necessary to calculate the probability of cracking in hardening concrete as reliable as possible in case of durability predictions. In this thesis, it is investigated how the accuracy of stress prediction in hardening concrete can be improved by a more detailed description of microstructural and transport phenomen. First it is described how hydration of cement and the microstructural development can be described. Two basically different models for cement hydration are explained, which introduced the possibility to couple results and information from the simulations with accurately property development of hardening concrete (strength, stiffness, porosity), however, the relation between visco-elastic behavior of hardening concrete and the microstructural development is still under development. Some history of development of visco-elastic models is presented, from which it turned out that there is a need for a step forward in understanding and explicit modeling of visco-elastic behavior of hardening concrete, by taking into account the microstructural development of hardening cement paste and moisture movements in order to make stress predictions in hardening concrete elements more accurate and reliable. The moisture state is modeled explicitly in the solidification theory by including hollow permeable shells in a microstructural hydration model. A hollow shell, or annular space, is defined as the zone between an anhydrous cement grain and the gel shell. This space is filled with water. The Shell Deformation Model (SDM) as developed in this thesis relates the time dependent deformations of a loaded cement paste to the transport of load bearing water inside the shell of cement gel to the capillary pores. The model assumes equal time dependent deformations for cement paste loaded in tension and compression. From the simulated results, it turned out that the fineness of cement does not influence the visco-elastic behavior of hardening concrete. In order to obtain reliable and accurate results from the proposed model, the moisture state in the capillary pores is considered accurately. For this purpose, moisture movements are modeled at microlevel. Due to the inhomogeneous distribution of cement particles in the cement paste between the aggregates, pressure differences exist between water in the pores of cement paste located near the aggregates and the water in the pores of cement paste in-between the aggregates. A model is developed which describe pressure driven moisture movements based on the conservation of volume. The influence of capillary water transport on visco-elastic behaviour of hardening cement paste is studied. Taking those movements into account, an influence up to 8% on the visco-elastic behaviour is observed. The differences are bigger for higher water/cement ratio’s and arise earlier for higher Blaine values of the used cement. As a first validation of the proposed model, two test series are performed. First test serie contains sustained loading experiments. The results are used to study the influence of the cement fineness and to study the influence of the load sign. Result of this program is that no influence on the visco-elastic behaviour can be ascribed to both parameters. Second test serie contains degree of restraint experiments to validate the Shell Deformation Model. For three mixtures with different cement finenesses, it turned out that stress predictions are predicted more accurate by using the proposed Shell Deformation Model (SDM) compared to a model previously proposed.