Objective:To explore the prognostic value of ferroptosis-related long non-coding RNA (lncRNA) in esophageal cancer.Methods:Based on the Cancer Genome Atlas (TCGA), the predictive model was constructed based on the differentially expressed ferroptosis-related lncRNAs in esophageal cancer.Results:Seventeen differentially expressed ferroptosis-related lncRNAs (AC108673.2, LINC00942, CASC8, AP003696.1, LINC02154, AC092969.1, AC245100.5, AC011379.1, TMEM161B-AS1, LINC00092, LINC01977, AC107308.1, LINC00261, LINC00592, AP000553.2, HOXC-AS1, Z93403.1) related to the prognosis of esophageal cancer were identified. Kaplan-Meier analysis showed that the high-risk lncRNA feature was associated with a poor prognosis of esophageal cancer ( P<0.01). The area under the curve of the lncRNA feature was 0.861 at 1-year, 0.828 at 2-year, 0.764 at 3-year; and it showed superiority over conventional clinical pathology features in predicting the prognosis of esophageal cancer. In addition, there were significant differences in immune cells between the low-risk and high-risk groups. The immune checkpoints such as TNFSF 9, CD70 and TNFRSF 25 were also different expression between the two risk groups. Conclusions:Ferroptosis-related lncRNA signature can precisely predict the prognosis of esophageal cancer and serve as therapeutic targets for esophageal cancer.