Blood leukocyte counts (e.g., eosinophil count) are important biomarkers for the onset, classification, and exacerbation of chronic obstructive pulmonary disease (COPD). The causal relationships between them are necessary for the development of COPD treatment strategy, but remain unclear. Here, we implement two-sample bi-directional univariable Mendelian Randomization (MR) and multivariable MR to investigate the causal re-lationships. Univariable MR find that elevated blood eosinophil count significantly increases the risk of COPD (odds ratio (OR) = 1.22, 95% confidence interval (CI): 1.14-1.30, P = 1.54 x 10- 09) and COPD-related hos-pitalization (OR = 1.44, 95% CI: 1.15-1.80, P = 1.36 x 10-03). Besides, it also significantly decreases the ratio of forced expiratory volume in the first second over forced vital capacity (FEV1/FVC ratio) (OR = 0.942, 95% CI: 0.914-0.971, P = 1.02 x 10-04). These findings are fully supported by multivariate MR results. Interestingly, univariable MR reveals a weak causal relationship between elevated blood eosinophil count and COPD risk in younger people (<65 years) (OR = 1.39, 95% CI: 1.10-1.75, P = 5.52 x 10-03), but not older individuals (OR = 1.20, 95% CI: 0.926-1.55, P = 0.17). Finally, reverse univariable MR reveals the onset of COPD and the decreased FEV1/FVC ratio both lead to increased blood neutrophil count (OR = 1.03, 95% CI: 1.01-1.05, P = 3.40 x 10-03 and OR = 0.947, 95% CI: 0.91-0.986, P = 8.75 x 10-03 respectively). In summary, this MR study demonstrates that high blood eosinophil count is an independent causal mediator of COPD risk, FEV1/FVC decline, and COPD-related hospitalization. The increase in neutrophil count is induced by COPD onset or FEV1/ FVC decline. This suggests eosinophil, but not neutrophil, may be used as a therapeutic target for preventing the onset and exacerbation of COPD and FEV1/FVC decline. Therefore, a non-neutrophil-targeted therapeutic strategy for neutrophilic COPD is required in the future.