Serial monitoring of peripheral blood lymphocyte subpopulations (PBLSs) counts might be useful in predicting post‐transplant opportunistic infection (OI) after kidney transplantation (KT). PBLSs were prospectively measured in 304 KT recipients at baseline and post‐transplant months 1 and 6. Areas under receiver operating characteristic curves were used to evaluate the accuracy of different subpopulations in predicting the occurrence of overall OI and, specifically, cytomegalovirus (CMV) disease. We separately analyzed patients not receiving (n = 164) or receiving (n = 140) antithymocyte globulin (ATG) as induction therapy. In the non‐ATG group, a CD8+ T‐cell count at month 1 <0.100 × 103 cells/μl had negative predictive values of 0.84 and 0.86 for the subsequent occurrence of overall OI and CMV disease, respectively. In the multivariate Cox model, a CD8+ T‐cell count <0.100 × 103 cells/μl was an independent risk factor for OI (adjusted hazard ratio: 3.55; P‐value = 0.002). In the ATG group, a CD4+ T‐cell count at month 1 <0.050 × 103 cells/μl showed negative predictive values of 0.92 for the subsequent occurrence of overall OI and CMV disease. PBLSs monitoring effectively identify KT recipients at low risk of OI, providing an opportunity for individualizing post‐transplant prophylaxis practices.