Aim The aim of the present study was to assess the validity and reliability of the Falls Efficacy Scale (FES) in Parkinson's disease (PD) patients in Serbia. Methods A cross‐sectional study was carried out at the Clinic for Neurology, between June 2011 and June 2012. A total of 201 consecutive PD outpatients were recruited. The inclusion criteria were: ability to walk independently for at least 10 m, ability to stand for at least 90 s and a Mini‐Mental State Examination score >24. The exclusion criteria were: the presence of other major neurological, psychiatric, visual, audio‐vestibular and orthopedic disturbances. The 10‐item FES was translated according to internationally‐accepted methodology. The internal reliability of the Serbian version of the FES was evaluated using Cronbach's alpha coefficient. Reproducibility of the FES was evaluated using the Spearman–Brown coefficient. To evaluate construct validity, an exploratory factor analysis (principal component analysis, varimax rotation) was carried out. Results The internal consistency of the Serbian version of the FES was 0.98. Age, duration of disease, Hoehn and Yahr stage, Unified Parkinson's Disease Rating Scale motor score, history of falls, and the Hamilton depression and anxiety scores were significantly correlated with the total FES score. On factor analysis, all 10 items were compact in a one‐factor cluster, with an explained variance of 85%. Spearman–Brown's correlation coefficient between the total scores was 0.99. Conclusions The psychometric characteristics of the Serbian version of the FES have excellent reliability and validity as an instrument for measuring the fear of falling in PD patients. Geriatr Gerontol Int 2013; ●●: ●●–●●.