Objective: The objective of this study was to investigate the relationship between the triglyceride-glucose (TyG) index, triglyceride-glucose-BMI (TyG-BMI) index, laboratory indices, and disease severity in patients with COVID-19. Methods: A retrospective analysis of COVID-19 patients treated at a tertiary hospital in Mianyang City, Sichuan Province, China, from 1 May to 31 May 2023 was performed. The patients were divided into two groups: 66 cases in the moderate group and 61 cases in the severe group. Additionally, 69 uninfected individuals from the medical examination center during the same period were selected as the control group. Spearman rank correlation was used to determine the correlation between the indices and COVID-19 severity. Multiple logistic regression analysis was performed to identify the factors affecting COVID-19 severity. ROC curves were constructed to assess the predictive value of the TyG and TyG-BMI indices for severe COVID-19. Results: There were significant differences in smoking and diabetes between the three groups (P < 0.05). The levels of ALT, AST, TyG index, and TyG-BMI index were higher in the severe group compared to the moderate and control groups, while the levels of ALB were lower in the severe group (P < 0.05). Correlation analysis showed that ALT, AST, TC, TG, HbA1c, TyG index, and TyG-BMI index were positively correlated with COVID-19 severity, while ALB was negatively correlated (P < 0.05). The multivariate logistic regression analysis revealed that AST, ALB, TyG index and TyG-BMI index were risk factors for moderate COVID-19, and smoking, AST, ALB, TyG index, and TyG-BMI index were risk factors for severe COVID-19. ROC curves demonstrated that the TyG index predicted an area under the curve (AUC) of 0.