Background: In China, the main treatment of esophageal squamous cell carcinoma (ESCC) is surgery combined with postoperative adjuvant chemoradiotherapy.The role of preoperative neoadjuvant chemoradiotherapy is not well established. We compared neoadjuvant chemoradiotherapy followed by surgery with surgery followed by adjuvant chemoradiotherapy in a Chinese ESCC population. Methods: We randomly assigned patients with resectable locally advanced tumors (T3-4N0-1M0, T1-2N1M0) to receive surgery and weekly administration of carboplatin (AUC=2) and paclitaxel (50 mg/m²) for 6 weeks and concurrent radiotherapy (50.4 Gy/28f, 5 days per week) at preoperative (the neoadjuvant group) or postoperative (the adjuvant group). Results: From April 2011 through December 2013, we enrolled 42 patients: 23 were randomly assigned to chemoradiotherapy followed by surgery, and 19 to surgery followed by adjuvant chemoradiotherapy. Among these 42 patients, the most common major hematologic toxic effects were leukopenia (9.5%), neutropenia (11.9%), thrombocytopenia (14.3%), and anaemia (16.6%); the most common major nonhematologic toxic effects were anorexia (14.3%), fatigue (11.9%), and cervical anastomotic fistula (19.1%). Complete resection with no tumor of the resection margins (R0) was achieved in 100% of patients in the neoadjuvant group versus 90.4% in the adjuvant group. A pathological complete response was achieved in 8 of 23 patients (34.8%) who underwent resection after chemoradiotherapy. Postoperative complications and treatment-related mortality were similar in the two groups. The disease free survival rate at 18 months was 78.7% in the neoadjuvant group as compared with 63.6% in the adjuvant group, which exceeded the goal of this study design. Conclusions: Our preliminary result suggests that, in patients with resectable locally advanced ESCC, there is a benefit tendency for the preoperative neoadjuvant chemoradiotherapy compared with postoperative adjuvant chemoradiotherapy. The regimen was associated with acceptable adverse-event rates. These trends warrant further study. Clinical trial information: ChiCTR-TRC-12002665.