This study explores the causal relationship between circulating immune cells and diabetic nephropathy (DN) progression using Mendelian randomization. We selected instrumental variables from genome-wide association study summary statistics, which included datasets of circulating leukocytes, flow cytometric subgroups of lymphocytes, and DN. Applying the standard Mendelian randomization approach and refining it with sensitivity analysis and horizontal pleiotropy examination, we found a significant causal association between increased eosinophil counts and a higher risk of DN (odds ratio: 1.21, 95% confidence interval: 1.07-1.37, P = .00027), which remained robust after further analysis. However, no causal link was observed between lymphocyte flow cytometric subgroups and DN. These findings suggest that elevated eosinophil counts may contribute to the progression of DN, potentially advancing to end-stage renal disease. The lack of a causal relationship between lymphocyte subtypes and DN underscores the need for larger scale genome-wide association study data for further validation. This research provides important causal evidence for considering eosinophil counts as a potential biomarker in DN diagnosis and management.