Objectives: Erythropoietin resistance index (ERI) seems to be associated with adverse outcomes in patients receiving chronic hemodialysis. Despite the role of erythropoietin in non-hematopoietic tissues, there is a lack of clinical studies showing the impact of body composition on responsiveness. The aim of this study was to examine the association of body composition with ERI using the longitudinal data in patients on chronic hemodialysis. Methods: This study included outpatients on chronic hemodialysis who received body composition analysis, which was performed every 6 months. The relationship between body composition and baseline ERI was evaluated in the multiple linear regression analysis, and the trend of ERI over time according to the changes in body composition was analyzed using a linear mixed effect model. Results: There were 65 (52.8%) men and 58 (47.2%) women, and they were followed up for 24 (5, 75) months. Subjects with higher mean ERI had the increased risks for all-cause death, cardiovascular events, and infection requiring hospitalization, compared to those with lower mean ERI.  ERI was negatively correlated to body mass index, percent body fat, arm circumference, arm muscle circumference, and visceral fat area. However, the ratio of extracellular water to total body water and skeletal muscle mass index was not related to ERI. The longitudinal trend of ERI was assessed according to the slope of body composition parameters over time. The increase in ERI was associated with the increase in the ratio of extracellular water to total body water, on the other hand, the trajectory of ERI was inversely proportional to the changes in arm circumference, arm muscle circumference, and visceral fat area over time. Conclusions: Erythropoiesis-stimulating agent responsiveness may be associated with body composition, especially with arm circumference, arm muscle circumference and visceral fat area. Therefore, body composition analysis can be useful to predict ERI in patients receiving chronic hemodialysis.