Background: Several studies have been performed to evaluate surrogate markers of long-term allograft function in renal transplant recipients. These include serum creatinine, estimated glomerular filtration rate (eGFR), slope of eGFR, and more recently eGFR variability. The aim of this study was to measure eGFR slope while assessing the variability of this slope and if high variability occurring at any time post-transplant was predictive of poorer long-term outcomes in a large cohort of kidney transplant recipients. Methods: Adult solitary kidney transplant recipients transplanted between July 1, 2005 and July 31, 2015 were included. The primary outcome was time to graft loss, defined as return to chronic dialysis, retransplant, or death. Secondary outcomes were death-censored graft loss and acute allograft rejection. Cox regression was utilized for primary and secondary outcomes. Multivariate logistic regression was used to determine baseline factors predictive of high eGFR variability. Results: A total of 1,543 patients were included in the analysis. The percentage of patients who experienced an eGFR coefficient of variation of <30% was 79.6% (1,229/1,543), while 20.4% (314/1,543) patients had high eGFR variability (≥30%). Patients with high eGFR variability tended to be younger, African-American and female. Those with higher eGFR variability, accounting for confounding and other eGFR measures (peak and slope), had significantly lower overall patient and graft survival. Conclusion: This study provides a novel analysis of the utility of eGFR variability in a large cohort. The clinical use of the slope of eGFR and eGFR variability may aid in predicting long-term graft outcomes and facilitate early patient discussions to change the trajectory of allograft function.