In order to foster the deployment of photovoltaics (PV) at the terawatt scale, the research for low-cost, earth-abundant and non-toxic PV materials, as well as cost-efficient fabrication processes, is intensified. Metal oxide semiconductors are today widely applied in PV as transparent conductors (transparent conductive oxide or TCOs) and functional interfacial layers. Some of these oxide semiconductors are even use as active solar absorbers, as they have appropriate bandgap to harvest a wide part of the solar spectrum. They are attractive materials as they can serve much functionality in a PV cell, chemically stable and a large number of them are non-toxic and abundant. Most importantly, they can also be fabricated by cost-effective, vacuumand solution-based fabrication methods, in the form of thin films and nanostructures. This thesis concerns the investigation of metal oxide semiconductor solar cells, based on heterojunction of the cuprous oxide (Cu2O) absorber and magnesium-doped zinc oxide (ZnMgO) buffer. We focus on the development of the single layers and the formation and engineering of the Cu2O/ZnMgO heterojunction, so as to optimize the solar cell power conversion efficiency. Cu2O, the most investigated metal oxide absorber, is a p-type semiconductor with a band gap of 2 eV, which makes it theoretically possible to yield efficiencies up to 20% for singlejunction solar cells (Shockley-Queisser limit). Cu2O is non-toxic, abundant and can be fabricated with non-vacuum deposition techniques, such as by electrochemical deposition (ECD). ECD is a solution-based fabrication process, in which the material usage and energy consumption is very low, while it is easily up-scalable to industrial-scale production. It is therefore an ideal fabrication method for low-cost, thin-film solar cells. A suitable n-type layer, forming the heterojunction with the p-Cu2O, is vital to obtain high power conversion efficiencies. Most important is the conduction band alignment at the p/n heterojunction. Zinc oxide (ZnO) is one of the most studied heterojunction partner for Cu2O. It is an intrinsic n-type semiconductor with a wide band gap of 3.4 eV, but its conduction band alignment with Cu2O is far from ideal. Doping the ZnO with Mg offers the possibility to modify the electronic band structure of the material. By increasing the Mg content, the bandgap of ZnMgO increases, while the crystal structure of ZnO is maintained. The increase of bandgap is accompanied with a smaller electron affinity, which improves the alignment of the conduction band with Cu2O. Further, to control the defect density at the Cu2O/ZnMgO interface, the atomic layer deposition (ALD) method is employed for the fabrication of the ZnMgO layer. ALD is a vacuum technique, widely applied in the semiconductor industry and of high-potential for thin-film PV. It is based on the sequential and self-limiting surface reactions to produce films of high quality and unsurpassed homogeneity even on 3-dimensional substrates. In the thesis we also sought to replace precious metals, used as solar cell electrodes in the literature, with lower-cost alternatives. The optimized solar cells in this thesis are precious-metal-free, almost exclusively composed of metal oxide semiconductor layers. For this, a highly conductive and reflective electrode, based on chromium (Cr) and indium-tin-oxide (ITO) was developed by sputtering. The ECD absorber deposition was optimized for the Cr/ITO electrode, in order to obtain void-free and large-grain Cu2O films. Finally, sputtered aluminum-doped ZnO (AZO) was used as front transparent contact to complete the solar cell. Best cells showed promising PV performance, with short circuit current density of 6.8 mA/cm2, open circuit voltage of 550 mV, fill factor of 45% and power conversion efficiency of 1.67%.