Infections associated with prosthetic devices involve an increased risk of persistence or relapse if the foreign body is not removed. Clinical and experimental observations indicate a lack of correlation between the minimal bactericidal concentration of a drug and its efficacy in device-related infections. Alternative in-vitro tests indicate that the efficacy of a drug experimental implant-associated infections depends on its activity on non-growing and adhering microorganisms. These data have been evaluated in a clinical pilot study. 12 patients with orthopedic device-related infections were treated with rifampin-containing regimens because this drug has excellent activity on adherent microorganisms. The success rate was 83% during a follow-up period of 2 years or until the removal of the device. Nevertheless, before considering conservative treatment of device-related infections, controlled clinical studies are needed which confirm the experimental results.