The study was conducted in bahía Chipana (21o S 71o W), located 170 km south of Iquique city at northern Chile, where extense Lessonia trabeculata Villouta and Santelices 1986 and Macrocytis integrifolia Bory 1826 kelp beds are inhabited by a diverse fish assemblage, composed of species which take advantage of the shelter provided by the habitat structure and Vásquez 1987, Godoy 2000, Angel and Ojeda 2001, Aguirre 2004). The first aim of the study was to examine whether the species composition, abundance, and spatial distribution of coastal fish are related to variations in the habitat structure. A second objective of the study was to examine whether trophic relations of coastal fish show differences between M. integrifolia and L. trabeculata kelp beds habitats. In situ experiments were carried out in order to analyze the fishes response to structural disturbance of their habitat (M. integrifolia, L. trabeculata). In a first experiment boulders colonized by M. integrifolia (n = 28/replicate) were transported into three replicated experimental areas (9 m2 each) located in the barren ground area. During the present and second experiment an artificial kelp bed was installed in a barren ground habitat. Three groups, each with 28 artificial plants homogeneously installed in an area of 9 m2 were set up. Other three natural areas of the same dimension were used as control. The fish abundance, composition and spatial distribution (rock, macroalgae, water column, sand) were recorded by SCUBA diving during three months. Initially the experiments was supposed to be repeated once or twice with some modifications depending on the preliminary results and the resistance of the experimental kelps to environmental conditions. Each artificial plant consisted of a plastic base, three stipes made of rubber hose and a terminal frond of polyethilen folium. The artificial plants were arranged in groups of seven on a square of fishing net, with its margins reforced by PVC pipeline. In the field the squares were fixed to the bottom by stones. The three experimental areas were set up during the 5/3/2008. Monitoring of experiments and comparative natural areas were done during 6/3/2008 and 17/6/2008. The results indicate a total of 6 fish species recorded during the experiments in the natural kelp, 4 of them (Isacia conceptionis, Mugiloides chilensis, Scartichthys gigas/viridis and Cheilodactylus variegatus) resident species. The artificial kelp was only visited by Scartichthys viridis/gigas and Isacia conceptionis, both not directly depending on the presence of kelp. The scarce results obtained by the experiment are probably a result of the plant design and its performance in streaming water, generating a too low canopy and scarce frond density that suggest design modification for eventual new experiments.