Objective: Anti-tuberculosis drug-induced liver injury (ATB-DILI) is a common adverse drug reaction during anti-tuberculosis treatment. Systemic inflammatory markers (NLR, MLR, PLR, and SII) have been shown to be independent predictors of many inflammatory and tumor-related diseases. However, the relationship between systemic inflammatory markers and ATB-DILI remains unknown. Our aim was to investigate the association between systemic inflammatory markers and ATB-DILI to examine the potential of systemic inflammatory markers in predicting ATB-DILI. Methods: A retrospective analysis was performed on 172 tuberculosis patients hospitalized in the tuberculosis ward of West China Hospital of Sichuan University. According to the occurrence of ATB-DILI, the patients were divided into ATB-DILI group and No ATB-DILI group. NLR, MLR, PLR, and SII values were calculated from the results of blood routine examination at admission. Binary logistic regression analysis was used to analyze the relationship between NLR, MLR, PLR and SII values and ATB-DILI. Results: Compared with No ATB-DILI patients, ATB-DILI patients had significantly lower levels of NLR, MLR, PLR and SII on admission, especially MLR. Among the systemic inflammatory markers, PLR had the strongest predictive ability for ATB-DILI, with an area under the curve (AUC) of 0.6769. Logistic regression analysis showed that MLR was an independent risk factor for ATB-DILI. Conclusion: Our study showed that baseline NLR, MLR, PLR and SII levels were associated with the occurrence of ATB-DILI. More clinical studies are needed to further verify the value of systemic inflammatory markers in ATB-DILI.