The aim of this study was to investigate the relationship among phase angle (PA), malnutrition, and prognosis in patients with gastrointestinal cancer.In total, 870 patients with gastrointestinal cancer were enrolled. Kaplan-Meier curves and Cox proportional hazards models were used to evaluate the association between PA and survival risk. Restricted cubic spline regression was used for flexibility analysis to explore sex-specific associations between PA and survival. Logistic regression was used to assess the relationships among PA, malnutrition, and cachexia.Low PA was closely associated with poor physical conditions, diminished quality of life, and malnutrition. Patients with low PA had a significantly worse prognosis than those with high PA (60.6% versus 72.8%; log-rank P < 0.001). PA was suitable for the prognostic assessment of patients with advanced-stage tumors. Regardless of sex, patients with lower PA showed significantly poorer survival rates. Cox proportional hazards models identified PA as an independent predictor of prognosis in patients with gastrointestinal cancer (hazard ratio (HR)=0.534; 95% confidence interval (CI)=0.409-0.696, P < 0.001). Subgroup analysis indicated that a high PA was an independent risk factor affecting the prognoses of patients with esophageal, liver, and intrahepatic bile duct cancers. Interestingly, variations in PA had a more significant prognostic effect on survival in men than in women. The logistic regression model confirmed that PA is a valuable indicator for assessing malnutrition and cachexia in patients with gastrointestinal cancer. Among all body composition indicators, PA demonstrated the highest accuracy for prognostic prediction.PA was identified as a robust predictor of malnutrition and poor prognosis in patients with gastrointestinal cancer.Copyright © 2024 Elsevier Inc. All rights reserved.