Different measures of rates of transfer of glucose during the peritoneal equilibrium test (PET), undertaken during peritoneal dialysis (PD) might provide additional information regarding a patient's risk of future cardiovascular mortality. This study aimed to characterize the heterogeneity of dialysate glucose (DG) response phenotypes during the PET and compare the cardiovascular mortality rates associated with the different phenotypes. Our cohort was derived from Henan peritoneal dialysis registry. A total of 3477 patients initiating PD in 2007 to 2014 had the DG measured at 0, 2-hour and 4-hour (D0, D2, and D4 respectively) during the PET for estimation of D2/D0 and D4/D0. Deaths mainly due to CVD within 2 years since the initiation of PD were defined as the outcome. Latent class mixed-effect models were fitted to identify distinct phenotypes of the DG response during the PET. Multivariable unconditional Logistic regression models with adjustment for cardiometabolic risk factors were used to compare the 2-year risk of cardiovascular mortality among patients in the different latent classes. Three distinct DG response phenotypes during the PET were identified. Those with consistently high D2/D0 and D4/D0 ratios had a 1.22 [95% confidence interval: 1.02, 1.35] excess risk of a cardiovascular death within 2 years of commencing PD compared with patients with the lowest D2/D0 ratio and decreased D4/D0 ratio after adjustment for cardiometabolic risk factors. Consistently elevated D2/D0 and D4/D0 ratios during the PET are associated with an increased risk of 2-year cardiovascular mortality independent of other cardiometabolic risk factors. In view of the potential bias due to unmeasured confounders (eg, Family history of cardiovascular diseases, and dietary patterns), this association should be further validated in other external cohorts.