Using data from the real world to solve clinical questions that cannot be answered using data from clinical trials is attracting more attention. Clinical outcomes for patients with esophageal cancer in a real-world setting might be different from data in randomized controlled trials. This study aimed to provide real world data on treatment and prognosis in Korean patients with esophageal cancer. This retrospective cancer cohort included newly diagnosed cases of esophageal cancer at 19 tertiary hospitals between January 1, 2005 and December 31, 2017. Cancer staging was defined according to the 7th edition of the American Joint Committee on Cancer criteria. We identified 6,354 patients with newly diagnosed esophageal cancer (mean age: 64.9 ± 9.0 years, 96.9% squamous cell carcinoma). The proportion of early esophageal cancer increased from 24.7% in 2005 to 37.2% in 2015 (p<0.001). Among all cases, surgery alone was 31.3%, followed by definitive concurrent chemoradiotherapy (CCRT) (27.0%), neoadjuvant therapy (12.4%), adjuvant therapy (11.1%), and endoscopic resection (5.8%). The 5-year overall survival rate was 45.7 ± 0.7%. Endoscopic resection provided similar median survival relative to surgery for stage Ia cases. Among stage II-III cases, definitive CCRT was associated with poorer survival than neoadjuvant or adjuvant therapy, although there was no survival difference between neoadjuvant and adjuvant therapy. Early esophageal cancer is gradually becoming more common and endoscopic resection provided similar long-term survival relative to surgery. Surgery with combined therapy provided better survival in locally advanced esophageal cancer, relative to definitive CCRT.