Backgroud: Traumatic brain injury (TBI) ranks among the leading causes of death worldwide. However, the association between hemoglobin variation (Delta Hb) and hospital mortality in TBI patients at high altitude remains uninvestigated. Method: This retrospective cohort study was conducted from January 2020 to March 2025 in the Tibetan Plateau region, enrolling 191 patients who resided at an average altitude of 3,000 m. Delta Hb (peak Hb-nadir Hb) during the hospitalization, related covariates and hospital mortality were collected. Backward stepwise multivariable logistic regression was used to select key variables. The non-linear relationship between Delta Hb and mortality was investigated using the multivariable fractional polynomial (MFP) method. The threshold effects of Delta Hb were explored through two-piecewise logistic regression models. Results: Logistic regression showed that Delta Hb was independently and significantly associated with hospital mortality (OR = 1.08, 95% CI: 1.02-1.15, P = 0.005) after adjustment for nadir Hb, diffuse axonal injury and GCS (Glasgow coma scale) score. A cubic non-linear relationship between Delta Hb and hospital mortality was revealed (P for non-linearity = 0.010), with an inflection point at 19.8 g/L. Additionally, an interaction effect between Delta Hb and GCS score was found (P = 0.035). Conclusions: In the clinical management of high-altitude TBI patients, our findings suggest that those with a Delta Hb > 20 g/L, and a substantially elevated Delta Hb and a low GCS score have an increased risk of mortality. A study investigating interventional strategies aimed at reducing Delta Hb in TBI patients is warranted.