This work lies at the borderline between surface science and magnetic materials. The aim is to make use of surface science processes, with a view to fabricating self-assembled nanostructures of interest for magnetic studies: ·Concerning fundamental research: study magnetization processes that occur in applied materials, but that cannot be analyzed quantitatively owing to the complexity of these materials. Self-assembled nanostructures are then used as model systems. ·The integration of self-organized nanostructures in devices is so far not conceivable, because such nanostructures are generally so thin, one or two atomic layers, that they do not retain a ferromagnetic behavior at room temperature, owing to their low dimensionality. Two growth processes have been developed that allow one to fabricate self-organized dots and stripes of height several nanometers, thus producing magnetic nanostructures functionnal at room temperature.