The Geriatric Nutritional Risk Index is a simple nutritional screening method, and this study aimed to investigate the association between the initial Geriatric Nutritional Risk Index and all-cause mortality in incident patients in the first year after the initiation of hemodialysis. This study is a retrospective cohort study and used the Korean Renal Data System database. Patients who were eligible for Geriatric Nutritional Risk Index assessment and underwent hemodialysis from January 2016 to December 2019 were included. The primary outcome was all-cause mortality, and outcome evaluation was performed in December 2020. A Cox proportional hazard model was used to analyze the association between the Geriatric Nutritional Risk Index and mortality. A total of 10,545 patients were included, and the mean age was 63.9 ± 3.7 years. The patients were divided into four groups by the quartile of the Geriatric Nutritional Risk Index with a mean value of 96.2 ± 8.2. During the study period, 545 (5.2%) deaths occurred. The surviving patients had higher Geriatric Nutritional Risk Index values than ones who died in the first year of hemodialysis initiation (96.6 ± 7.5 vs. 88.2 ± 9.3, p < 0.001). Quartile 1 (Geriatric Nutritional Risk Index < 91.8) showed a significantly increased risk of all-cause (Hazard Ratio: 2.56; 95% Confidence Interval: 2.13-3.09; p < 0.001) and cardiovascular mortality (Hazard Ratio: 22.29; 95% Confidence Interval: 1.71- 3.08; p < 0.001) at the first year in comparison with Quartile 4 (Geriatric Nutritional Risk Index ≥ 101.3). In areas under the receiver-operating characteristic curves of all-cause mortality, the Geriatric Nutritional Risk Index model improved predictive values, compared to the baseline model. The area with the Geriatric Nutritional Risk Index model was significantly higher than the one with a model including albumin or body mass index (p < 0.001). These findings suggest that a low Geriatric Nutritional Risk Index (<91.8) is associated with first-year all-cause and cardiovascular mortality in patients who start hemodialysis and may be a useful and reproducible tool for assessing prognoses in this population.