Variceal hemorrhage may cause high rebleeding and mortality rates. Preventing the first episode of variceal bleeding is mandatory in patients with high-risk esophageal varices (EV). This study aimed to identify factors that predict the recurrence of EV after endoscopic treatment (ET), and to develop a reasonable therapeutic strategy for EV in cirrhosis. From January 2012 to December 2014, 45 patients with cirrhosis and high-risk EV underwent ET, including sclerotherapy and/or ligation. Statistical analyses identified factors associated with the recurrence of EV after ET, and the Kaplan-Meier method determined the cumulative variceal recurrence rates. The 1-, 2-, and 3-year cumulative posttreatment recurrence rates for EV were 13.3%, 29.5%, and 32.2%, respectively. No significant differences were evident between the patients with and without variceal recurrences at 1-year posttreatment. The multivariate regression analyses identified a history of partial splenic embolization (PSE) and the pretreatment Child-Pugh classification as independent predictors of variceal recurrences at 2 years (p < 0.05) and 3 years (p < 0.05) posttreatment. While EV did not recur after ET and splenic artery embolization in cases with Child-Pugh class A, the overall posttreatment variceal recurrence rates were 0% and 66.7% when PSE was performed before and after ET, respectively, in those with Child-Pugh class B or C. Splenic artery embolization significantly reduced the hepatic venous pressure gradient and markedly lowered the Child-Pugh score in 15 patients. Adjunctive PSE and pretreatment Child-Pugh class A could be independently associated with reduced cumulative recurrence rates of EV post-ET. From the perspectives of portal hemodynamics and hepatic function, splenic artery embolization before or after ET could prevent posttreatment variceal recurrence in patients with Child-Pugh class A, and PSE before ET could achieve the long-term eradication of EV following ET in those with Child-Pugh class B or C.