The aim of this study was to estimate the prevalence of rheumatoid arthritis (RA) in Serbia, using the European EULAR project methodology. In a detection phase, a previously translated and validated telephone Questionnaire was used by lay interviewers on 6,213 randomly selected telephone numbers representing urban population from four Serbian towns: Belgrade (north), Cacak, Uzice and Krusevac (south). Patients with suspected RA were called again by a rheumatologist. For patients with self-reported diagnosis and positive symptoms, patient's rheumatologist was contacted to confirm diagnosis; a complete rheumatologist examination was scheduled for those with positive symptoms only. Prevalence estimates were standardised for age and sex in relation to Serbian population (census 2002) and further to French population, according to EULAR project methodology. The response rate was 63.6 % (3,950 respondents). The rheumatologist called 571 people, among whom 23 RA cases were confirmed (21 diagnosed previously and 2 newly diagnosed during the examination). The prevalence was 0.16 % (95 % confidence interval CI 0.01-0.32) for men and 0.51 % (95 % CI 0.26-0.76) for women; a female-to-male ratio 3.18. The overall Serbian standardised prevalence was 0.35 % (95 % CI 0.18-0.52); when standardised on French population 0.34 % (95 % CI 0.17-0.51). The highest age-specific rate was in the 65-74-year age band. The EULAR prevalence study, conducted with similar methodology and design, showed that RA prevalence estimates in Serbia (0.34 %) were in accordance with France (0.31 %), but lower than in Lithuania (0.55 %).