This study aimed to evaluate the causal effects of different immune cells on heart failure (HF) using Mendelian randomization (MR). Datasets for immune cell phenotypes and HF were obtained from European Bioinformatics Institute and FinnGen. Then, single nucleotide polymorphisms were screened according to the basic assumptions of MR. Subsequently, inverse variance weighted was used as primary tool for MR analysis, and Cochran Q and leave-one-out analyses were used to assess heterogeneity and robustness, respectively. MR analysis showed that cluster of differentiation (CD) 66b++ myeloid cell absolute count (AC) (odds ratio [OR] 1.043, 95% confidence interval [CI] 1.001–1.088, P = .045), human leukocyte antigen D-related on CD14– CD16+ monocyte (OR 1.030, 95% CI 1.005–1.056, P = .019), IgD on unsw mem (OR 1.046, 95% CI 1.015–1.078, P = .003), CD4 on CD4+ (OR 1.039, 95% CI 1.009–1.070, P = .011), CD24 on IgD+ CD38– (OR 1.026, 95% CI 1.000–1.052, P = .046), CD20 on CD24 + CD27+ (OR 1.032, 95% CI 1.003–1.061, P = .029), CD19 on CD20– (OR 1.037, 95% CI 1.005–1.071, P = .023), CD62L– CD86 + myeloid dendritic cell %DC (OR 1.032, 95% CI 1.004–1.061, P = .027), human leukocyte antigen D-related + CD4 + AC (OR 1.037, 95% CI 1.003–1.072, P = .032), and effector memory CD8br AC (OR 1.048, 95% CI 1.021–1.076, P < .001) were associated with increased genetic susceptibility to HF. Cochran Q and sensitivity analyses showed that the results had no heterogeneity and were robust. This MR analysis revealed 10 immune cell phenotypes associated with increased genetic susceptibility to HF. These findings provide new directions for understanding the pathogenesis of HF and developing novel therapies.