Objectives: Goals of epidemiological studies are the description of the measures of frequency of diseases, the attempt to clarify possible etiopathogenic mechanisms, and the provision of data to support health policy decisions. To increase the familiarity of rheumatologists toward epidemiology, we describe the methodology used in a prevalence study of musculoskeletal complaints performed in Chiavari, Italy. Methods: A questionnaire, originally developed by the Epidemiology Unit of the Arthitis Research Council in Manchester, UK, to investigate the prevalence of rheumatoid arthritis, was used after translation and validation. 4456 subjects aged 16 years or more listed in four general practices were invited to participate in the study and to fill the ARC questionnaire. The 3294 responders reported a) any past occurrence of joint swelling lasting more than four weeks and the distribution of the swollen joints on a mannequin; b) any joint pain lasting more than four weeks; c) current joint pain or swelling; d) morning stiffness; e) whether they had been previously told by a doctor they had arthritis. Results: Four steps were necessary to obtain a 74% response, i.e. direct contact, two mailings and a phone interview. The performance of the different questions was good. The prevalence of the most common conditions among patients answering positively to the questions regarding morning stiffness and symmetrical swelling of joints was as follows: osteoarthritis 2.60%, fibromyalgia 1.30%, carpal tunnel syndrome 1.14%, rheumatoid arthritis 0.31%, and psoriatic arthritis 0.10%. Conclusions: Methodological issues regarding the selection of the population and sample to study, the development of a questionnaire, and the problems in obtaining valid informations are discussed