The Geriatric Nutritional Risk Index (GNRI) is a valuable simplified tool to predict mortality. However, the results of previous studies are inconsistent and controversial. To summarize the evidence regarding the association of GNRI levels with the risk of all-cause and cardiovascular (CV) mortality, we conducted this meta-analysis. Relevant studies were identified through a systematic electronic literature search. We estimated combined hazard ratios (HRs) to assess the association between GNRI and the risk of mortality by using a meta-analysis method. The Cochrane Q test and the inconsistency statistic were used to assess the between-study heterogeneity. Subgroup analysis and sensitivity analysis were performed. Twenty-six observational studies involving 17 097 participants were identified in this meta-analysis. With the highest category used as the reference group, the lowest-category GNRI was significantly associated with an increased risk of all-cause (HR: 1.32, 95% confidence interval: 1.22-1.43) and CV (HR = 2.10, 95% confidence interval: 1.72-2.57) mortality. Subgroup analyses based on the participant ethnicity, age, and the duration of the follow-up period did not substantially change the main results. In summary, a lower GNRI is associated with an elevated risk of both all-cause and CV mortality. Given the significant heterogeneity among the included studies, further investigations with larger sample sizes are required to confirm the value of the GNRI in predicting mortality and to explore the combined effects of malnutrition and mortality.