Background: Epicardial adipose tissue (EAT) has been related to increased cardiovascular risk in chronic kidney disease patients. However, prospective studies of EAT thickness in prediction of cardiovascular events in CKD patients are lacking. Moreover, there are inconsistencies in literature regarding cut-off of EAT thickness, standard technique and phase of measurement. Objectives: This study was undertaken to compare systolic and diastolic EAT thickness in prediction of CV events in non-dialysis dependent CKD patients. Methods: In this prospective, observational study, transthoracic echocardiography (TTE) was used to assess systolic and diastolic EAT thickness in 210 consecutive non-dialysis dependent CKD patients and followed up for at least one year for pre-defined end-points. Results: The mean systolic and diastolic EAT thickness in the CKD group (5.6±1.2mm and 4.2±1.1mm) was significantly higher than the non-CKD participants (4.3±1.0mm and 3.1±1.1mm), both P<0.001. Interclass correlation coefficient (ICC) agreement on measurements were 0.93 (95% CI: 0.79-0.98) for systolic EAT and 0.91 (95% CI: 0.74-0.97) for diastolic EAT. On multivariate linear regression analysis, only e-GFR remained as independent predictor of both systolic and diastolic EAT thickness. Receiver operating characteristics (ROC) analysis showed that diastolic EAT thickness of 5mm and systolic EAT thickness of 3.8mm had similar sensitivity (88% versus 87%, respectively) and specificity (72% versus 74%, respectively) to predict CV events in CKD patients. Conclusion: Both systolic and diastolic EAT thickness are significantly increased in CKD patients and can be used in CV risk stratification with similar sensitivity and specificity albeit with different cut-offs.