Human amniotic fluid cells have been discovered as a promising cell source for cardiovascular tissue engineering and regeneration as they can be successfully used for the in vitro fabrication of autologous constructs before birth. However, prior to their definite clinical implementation, a thorough in-vivo assessment of these cells and the deriving constructs seems indispensable in order to evaluate the principal feasibility of using this cell source for cell-based therapies. In the end, also the evaluation of safety aspects regarding this cell source seems crucial. Therefore, different human and animal-derived fetal cells were isolated and characterized as part of the presented thesis. They were expanded under defined conditions and their ability for the generation of cardiovascular tissue engineered constructs was assessed in vitro. For the in-vivo application, a novel cell seeding and minimally invasive implantation method for heart valves was established in a nonhuman primate model and the mode of fetal valve delivery was assessed as part of a pilot investigation in the ovine system. After proving principal feasibility, autologous amniotic fluid cells were isolated and seeded onto biodegradable composite starter matrices. Next, the valves were crimped and delivered prenatally into the pulmonary position of ovine fetuses and followed until birth. The presented studies of this thesis demonstrate the principal feasibility of using different human and ovine fetal cells for the in vitro fabrication of cardiovascular tissue engineered constructs. The studies also show the technical feasibility of using mononuclear cell-seeded valve-shaped matrices as heart valve replacements for several weeks in-vivo, and finally they show the combination of these techniques resulting in the implantation of amniotic fluid stem cell-based heart valve constructs into the fetal sheep model as a possible next generation therapy of congenital heart valve disease.