Low-salinity water flooding is a promising technique for the improvement of oil recovery. Despite intensive research, there are controversies about the pore-scale mechanisms behind the low-salinity effect (LSE). One of the proposed LSE mechanisms is wettability alteration, which is reflected in a decrease in contact angle. However, it is not known how a variation in contact angle affects the macroscopic transport properties such as capillary pressure and relative permeability. The effect of contact angle variation on the transport properties is investigated by simulating capillary dominated, two-phase flow with a quasi-static pore-scale modeling method. The simulations are performed with real rock samples and, therefore, the effects of heterogeneities in the rock are incorporated. We distinguish three levels of analysis: single pore throat, bundle of tubes and pore network model. The first two levels of analysis comprise microscopic effects of the contact angle, whereas the pore network also includes macroscopic effects due to the connectivity characteristics of the network. For this reason, percolation theory is used to enhance the understanding of the effect of the connectivity on the flow properties. The pore-scale fluid configuration is influenced by the wettability of the pore walls. However, previous research has not related the conventional wettability definition with the definition from an engineering perspective, which is based on the capillary pressure curve. Therefore, to include wettability effects on the transport properties, the definition of "pore wettability" has been revised such that the curvature of the fluid-fluid interface is consistent with the capillary pressure. As a result, the pore wettability is not solely dependent on the contact angle, but also on the pore geometry. The geometry now plays a crucial role, which implies that it is of high importance that the extraction algorithm provides accurate pore geometries. Modeling results reveal that the contact angle has a negative effect on transport properties in perfectly homogeneous media. However, real rock representations always have a certain degree of heterogeneity. Heterogeneity negatively affects the transport properties, because it decreases the connectivity and thus the conductivity of the network. In presence of permeability contrasts, variation in contact angle influences the pore filling sequence, which affects the transport properties. Snap-off displacements suppress the effect of permeability contrasts, which means that regards oil recovery water-wet systems are favoured over oil-wet systems. In contrast, in case of heterogeneity at pore-level, snap-off leads to oil entrapment and has a negative effect on the residual oil saturation. This implies that oil recovery improves at increasing oil-wetness. For this reason, maximum oil recovery is obtained under intermediate-wet conditions. Since heterogeneities play an important role in snap-off, the optimum wetting state is dependent on the network topology. When aging is taken into account, an opposite flow behaviour is observed. Transport properties are positively - instead of negatively - affected by an increasing contact angle. These conflicting results might be an explanation for the inconsistent experimental results. Since capillary dominated flow neglects differences in displacement rates, further research is required to investigate the effect of contact angle variation in the presence of dynamic forces. Furthermore, a dynamic model would allow mobilization of the trapped oil phase, which enhances the accuracy of the oil recovery predictions. The size of the oil clusters might provide an indication on the obilization of the oil entrapment.