Resumo. BACKGROUND: Falls are an impactful problem for people with Parkinson's disease (PwPD), impairing the functional capacity and quality of life. Although some predictors of fall risk have been evidenced, the investigation of functional capacity components individual and combined is needed to a more detailed approach to identify fall risk prediction. AIM: To verify the ability of individual and combined functional capacity components in predicting fall risk in PwPD. METHOD: Ninety-six individuals participated in this cohort study. Functional capacity was assessed by Berg Balance Scale, Timed Up and Go, Six-Minute Walk, and Sit-to-Stand tests. The participants who fell at least once during the prospective follow-up of 12 months were considered fallers. A trained evaluator registered the fall’s occurrence weekly. Mann-Whitney U, t-tests, and chi-square tests were used to compare fallers versus non-fallers. Receiver Operating Characteristics curves of functional capacity components (individual and combined) were performed to predict fall risk. RESULTS: Thirty-six (37.5%) PwPD were classified as fallers, being recorded 56 falls, and 60 (62.5%) as non-fallers. There are no differences in functional capacity components between groups. The combined functional capacity components were not able to predict fall risk in PwPD (area under the curve = 0.52; p = 0.77). These components were also unable to predict fall risk when analyzed individually. CONCLUSION: Our results indicated that functional capacity components, individual and combined, are not good predictors of fall risk in PwPD. The multifactorial characteristics of falls and the falls classification might affect the fall risk prediction and should be further investigated.