Prostate cancer (PCa), a severe health burden for males, accounts for the second frequent cancer and fifth tumor specific death cancer around the world. Several studies on tumor‐infiltrating immune cells (TIICs) have shown inconsistent and controversial results to PCa. We downloaded a gene expression matrix and clinical information from TCGA, and CIBERSORT was used to identify the proportion of TIICs. Wilcoxon's Sign Rank Test evaluated different gene expression levels in PCa and normal tissues. Kaplan‐Meier curves were used to evaluate the associations of TIICs and recurrence‐free survival (RFS). Finally, based on the preset
P‐value
of .05, the distribution of TIICs in 73 PCa tissues and 11 normal tissues was illustrated. Activated CD4
+
T cells and M0 macrophages account for a high proportion in PCa tissues, while neutrophils and monocytes were found at a high density in normal tissues. Further results showed that the density of plasma cells, Treg cells and resting mast cells were associated with advanced PCa. Additionally, M2 macrophages affected the RFS of PCa patients, and AR was also involved. In the current study, we first evaluated the immune infiltration among PCa and revealed that M2 macrophages could predict the prognosis of PCa patients. Meanwhile, based on the LASSO regression analysis, we established a novel nomogram to assess the recurrence risk of PCa based on immune cell proportions and clinical features.