Objective: To detect the immune indicators before induction therapy for lupus nephritis (LN), and to identify the indicators with predictive value for the initial infection in LN patients during induction therapy, so as to provide guidance for the screening of high-risk patients with infection and preventive anti-infection. Methods: A total of 82 patients with active type Ⅲ/Ⅳ LN from January 2022 to December 2024 were selected as the research subjects. Peripheral blood lymphocytes, T lymphocyte subsets and B lymphocyte subsets were all detected before induction therapy. According to whether infection occurred in the early stage of induction therapy, they were divided into the infection group (n=39) and the non-infection group (n=43). The demographic, pathological classification, induction treatment regimens, dose of mycophenolate mofetil (MMF), and baseline immune indicators of the two groups of patients were collected for analysis. Results: There were no statistically significant differences in age, gender, body mass index, pathological classification, and induction treatment regimens between the infected group and the non-infected group (P>0.05), while the difference in MMF dose was statistically significant (P<0.05). There was no statistically significant difference in the number of CD3+, CD4+ and CD8+ cells, Th1 cells, Th2 cells, Treg cells and CD19+B between the infection group and the non-infection group (P>0.05). However, the differences in lymphocytes, helper T cell 17 (Th17), the Th1/Th2 ratio, and the Th17/Treg ratio were statistically significant (P<0.05). Multivariate Logistic regression analysis showed that the dose of MMF, the Th1/Th2 ratio, and the Th17/Treg ratio were independent influencing factors for the occurrence of infection (P<0.05). The dose of MMF, the ratio of Th1/Th2, and the ratio of Th17/Treg have moderately reliable predictive values for infection (AUC ranging from 0.7~ 0.9). The combined predictive value of the three indicators is high (AUC >0.9), and the sensitivity and specificity are 93.75% and 91.55%, respectively. Conclusion: Compared with uninfected individuals, the baseline Th1/Th2 ratio and Th17/Treg ratio imbalance in patients with infection in the early stage of induction therapy is more significant; Before induction therapy, the infection risk of LN patients can be comprehensively judged by combining the Th1/Th2 ratio, Th17/Treg ratio, and the proposed dose of MMF, etc.