Objective: To evaluate the influence of the longterm survival in pT3 lower segment esophageal squamous carcinoma after radical surgery by selective arterial infusion chemotherapy. Methods: One hundred and eight (T3N0M0 56 and T3N1M0 52) of the 190 patients with pathologic pT3 lower segment esophageal squamous carcinoma, confirmed by consecutive radical surgery from January 1990 to January 1999, were divided into the experimental group (surgery + intra-arterial infusion), and the rest were divided into the control group (surgery alone, T3N0M0 48 and T3N1M0 34). The experimental group was subgrouped into 56 cases (T3N0M0 26 and T3N1M0 30) with less than 3 periods 3 times of therapy, and 52 cases (T3N0M0 30 and T3N1M0 22) with more than 3 times of therapy (including 3 times). We adopted one to seven courses of selective arterial infusion chemotherapy with cisplatin (80 mg/m 2 of body-surface area) and fluorouracil (800 mg/m 2) with or without epirubicin 3 or 4 weeks after operation. The interval for each periods was 3 to 4 weeks. All cases were followed-up for more than 5 years. The survival rate was calculated by the method of Kaplan-Meier and the survival difference between the patients with and without selective arterial infusion chemotherapy was compared with log-rank test. Prognostic variables were enterd into a Cox regression model controlling for age, site, lymph node status, and treatment received. Results: The overall suevival rate had no significant difference between the experimental groups and control group in this experimental, but there was a better suevival rate in receiving more than 3 courses of chemotherapy (including 3). Conclusion: Transcatheter arterial infusion chemotherapy is a safe and effective method, postoperative selective arterial infusion chemotherapy can prolong the survival rate in patients with esophageal squamous carcinoma, who have previously underwent the radical surgery. However, further investigation will be needed in future.