The use of nanomaterials has been shown to be promising in potential applications such as magnetic storage devices, nano-optical devices and sensors, amongst others. Such applications require the disposition of nanoparticles, of a wide variety of materials and with a narrow size distribution, into large and ordered arrays. With this aim, it has been shown that by modifying the forces present in self-assembly processes of colloidal particles the fabrication of ordered nanostructures over large surface areas is possible, more precisely for particle systems with sizes smaller than 100nm. In the present work, various colloidal suspensions were used to fabricate self-assembled nanostructures. Silica 75, 45 and 20nm, gold 50, 30 and 15nm, iron oxide 10nm and zinc sulphide 5nm particles as aqueous suspensions were arranged on various substrates. Prior to self-assembly, these nanoparticles were characterized in their suspending medium. Zeta-potentials, and more generally colloidal stability, have been measured for various ionic strengths and pHs. Size distribution measurements using four different methods have been performed on these particles. Investigations have shown that difficulties are encountered when analyzing amorphous and isolating particles smaller than 30nm. The surface of the silica particles has been studied in more details, since abnormal colloidal stability was detected compared to that predicted by classical models for colloidal stability, especially close to the isoelectric point. This study revealed that a surface hairy layer was very likely to reinforce the repulsive interaction between particles, as a steric effect. This hairy layer probably derives from the synthesis process employed to produce the silica particles, and may be composed of silanol groups and/or polysilicic acid chains. Three-dimension colloidal crystals with ordering lengths of tens of micrometers have been obtained by drying, under controlled temperature and relative humidity, a highly (0.36g/ml – ∼15 % volume) concentrated 75nm silica suspension on a flat substrate in a so-called Teflon ring cell, or by using dip-coating. As a result, a thick cracked film was produced in which the particles were ordered in three dimensions, as a colloidal crystal. The ordering length, which could reach several tens of micrometers, was also present for 45nm silica particles but was lost when using 20nm particles, most probably due to polydispersity and colloidal stability. The ordering of the particles in their suspending fluid close to the end of drying, influenced by the repulsive forces was believed to be at the origin of this phenomenon. Investigations of the drying process were carried out using optical microscopy, spectrophotometry analysis and weight loss measurements. It was shown that this drying process could be related to the one observed in sol-gel science, and the stress induced in the film during drying of the suspending medium could be calculated using existing models. Capillary forces were compared to particle-particle dispersion forces to again confirm the probable presence of a hairy layer at the silica particles' surface. Further investigations were also pursued to check the in-situ ordering, but without definite confirmation of the effect. Using the Teflon ring cell or dip-coating and working with dilute suspensions (3.6·10-4g/ml – ∼2·10-2 % volume) of 75nm silica particles allowed the fabrication of particle monolayers for which pH and thus colloidal stability played an important role. Short-range ordered monolayers were obtained at pH 6 and 10, while agglomerates were observed at pH 2, i.e. close to the isoelectric point of the silica particles. The driving forces encountered during the formation of 2D and self-assembled films could be described using models of capillary forces, particle-particle and particle-substrate interactions. The particle concentration, the dip-coating speed and the wettability of the substrate were also investigated more in detail. Finally, monolayers of 10nm superparamagnetic particles, stabilised by electrostatic or steric effects, and obtained on silicon wafer were investigated by AFM. This shows how the presence of a polymer in suspension could help the deposition of particles on an atomically flat, non-wetting surface. These experimental and theoretical descriptions of the self-assembly process of 3D and 2D nanostructures were employed then to set the range of parameters required for exploring the next step: the guided self-assembly process. Dip-coating combined with topographically nanopatterned substrates, prepared by Extreme-UV Interference Lithography (EUV-IL), allowed the production of micrometers-long chains of silica and gold particles with diameters between 50 and 15nm. The structure and quality was shown to strongly depend on the various experimental parameters, as in the case of the 2D nanostructures preparation. The particle concentration of the suspension, pH, ionic strength and the removal speed of the substrate out of the suspension were shown to be the most relevant parameters on the way to obtain long lines with few defects. Similar dip-coating experiments using substrates patterned with holes lead to ordered arrays of single dots of silica and gold, and ordered clusters of iron oxide and zinc sulphide over tens of micrometers. As a final step in this study, some physical properties of the obtained nanostructures were investigated. Surface plasmon resonance measurements performed on gold particle chains using UV/vis spectroscopy. These measurement showed very coherent results compared to theoretical simulation and other works. Cathodoluminescence spectra were also obtained using zinc sulphide nanoparticle arrays, and conductivity measurements were attempted on gold particle chains. Finally, a feasibility study on the colloidal lithography process was performed with more or less success depending on the employed method.