<p>Cancer-specific mortality in esophageal squamous cell carcinoma (ESCC) is commonly estimated at the time of diagnosis using static survival models. However, as an increasing proportion of patients survive beyond the initial high-risk period, residual cancer-related risk may evolve over time. Using population-based data from 11,806 patients diagnosed with ESCC between 2004 and 2015 in the Surveillance, Epidemiology, and End Results (SEER) database, we investigated survivorship-dependent changes in cancer-specific mortality within a competing-risk framework. A baseline Fine-Gray competing-risk model was constructed to estimate cancer-specific mortality, and landmark-based dynamic analyses were performed at 12, 24, and 36 months after diagnosis. At diagnosis, the estimated cumulative incidence of cancer-specific mortality reached 66.9% at 60 months. In contrast, the estimated probability of cancer-specific death within the subsequent 60 months declined from 40.8% among 12-month survivors to 15.6% among 36-month survivors, suggesting substantial attenuation of residual cancer-specific risk with increasing survivorship duration. Dynamic discrimination and standardized net benefit in landmark-based dynamic analyses varied across survivorship stages, reflecting shifts in residual risk distribution over follow-up. These findings indicate that cancer-specific mortality in ESCC is strongly dependent on survivorship duration. Landmark-based dynamic assessment provides complementary insights to baseline estimates and supports survivorship-aware interpretation of long-term cancer-specific risk.</p>