The adsorption of an anionic surfactant on a variable-charge surface has been studied by comparing results obtained for sodium p-3-nonylbenzenesulfonate (SNBS) on rutile with theoretical predictions based on the self-consistent field lattice theory (SCFA) by Scheutjens, Fleer, and Leermakers. Both the measured and the calculated isotherms show strong cooperative effects at low coverages and have shapes which are characteristic for the adsorption of anionic surfactants on metal oxides. The experimental and calculated adsorption isotherms at different salt concentrations have a common intersection point (cip) corresponding to surface charge neutralization. Before the cip the surfactants are adsorbed with their head groups in contact with the surface. After the cip bilayer formation starts. Surfactant ions effectively screen the surface charge, so that this charge increases upon surfactant adsorption. From the surface charge the minimal amount of surfactant adsorbed 'head-on" can be estimated. This amount depends mainly on the surface potential, which is fixed by the pH, and hardly on the concentration of indifferent electrolyte. On the other hand, the amount of surfactant at the solution side of the bilayer mainly depends on the salt concentration and hardly on the surface potential. Due to the different charge screening mechanisms at the inner and the outer sides of the adsorbed layer, the layers are asymmetrical.