In the grayscale masked stereolithography 3D-printing process, a liquid polymer resin is selectively exposed to UV light in a layer-by-layer fashion. By varying the local grayscale value displayed on a projection LCD mask situated below the resin tank, the light exposure of each point can be adjusted. The exposure dose determines where the material is cured or remains liquid and affects the material properties, such that the local stiffness of the material increases with higher exposure. At this point, light propagation effects play a decisive role, leading to overcuring, i.e., a mismatch between the original input geometry and the print results. A simulation of the printing process including these effects enables an accurate prediction of the geometry and properties of the printed object. It can be used for optimization, e.g., to mitigate the overcuring effect. The ability of additive manufacturing techniques to print complex geometries provides the possibility to manufacture the designs generated by computational optimization methods, such as Topology optimization (TO). For example, TO can be used to maximize the stiffness of a structure with fixed amount of material, to obtain optimal designs for compliant mechanisms, or to generate microstructures. At this point, it can be important to include a simulation of the printing process in the TO, in order to take into account the characteristics of the print process and obtain designs that can directly be additively manufactured with high accuracy. The authors investigate how the aforementioned process simulation for stereolithography can be integrated into the TO. The grayscale input masks are used as design variables, such that the optimization result can directly be printed. Further, TO of functionally graded material designs can be implemented, based on the dependency of the material properties on the exposure dose.