This paper presents a case study of technical and economic optimization of the periodicity of predetermined preventive maintenance actions carried out on a repairable industrial system from an EDF electric power plant. This analysis is conducted with the MARS software tool (MARS for "maintenance assessment of repairable systems"), developed jointly by Grenoble University and EDF R&D. The innovative aspect of this work lies in the integrated approach that is used, combining two steps. A first estimation step retrospectively assesses maintenance effect on system reliability. A second simulation step predicts the behavior of the maintained system over the time period set as an objective by the operator. The different stages of the case study are described in detail with elaborated considerations about optimization of the periodicity of preventive maintenance.