The study aimed to examine whether omission of 5-fluorouracil (5-FU)-containing chemotherapy alters pathological complete response rates in patients receiving trimodality therapy for locally advanced esophageal cancer. A total of 159 patients were identified. One hundred twenty-nine patients received platinum/5-FU concurrently with radiotherapy, and 30 received taxane/platinum-containing chemoradiotherapy prior to esophagectomy. Patients were staged using the 2002 American Joint Committee on Cancer staging system. Patients were matched between chemotherapeutic groups, with no significant demographic or clinical differences other than T stage (14% T2 in the 5-FU group; no T2 in the platinum/taxane group) and radiotherapy technique (8.5% received intensity-modulated radiotherapy in the 5-FU group; 60% in the platinum/taxane group). Pathological complete response rates for 5-FU and platinum/taxane-based groups were not significantly different (45% and 30%, respectively; P = 0.1548). Five-year overall survival and progression-free survival were not statistically different between the two groups. Significant predictors of pathological complete response included N stage (56% N0 and 33% N1; P = 0.0083), histology (37% adenocarcinoma and 59% squamous cell; P = 0.0123), tumor location (39% distal and 59% proximal/mid; P = 0.048), gastroesophageal junction involvement (33% involved and 55% uninvolved; P = 0.005), and radiotherapy end-to-surgery interval (50% < 55 days and 34% ≥ 55 days; P = 0.04). Grades 3-4 hematological toxicity was higher in the 5-FU group (36%) than in the paclitaxel-containing therapy group (17%; P = 0.0484). Use of paclitaxel-containing chemoradiotherapy did not result in inferior pathological complete response, overall survival, or progression-free survival rates, and resulted in less hematological toxicity than 5-FU treatment.