BACKGROUND The incidence of prostate cancer (PCa) is on the rise in China. The risk level of patients with PCa is associated with disease-free survival rate at 10 years after radical prostatectomy. Predicting prognosis in advance according to the degree of risk can provide a reference for patients, especially treatment options and postoperative adjuvant treatment measures for high-risk/extremely high-risk patients. AIM To explore the predictive value of the prognostic nutritional index (PNI) for biological recurrence in Chinese patients with high/extremely high-risk PCa after radical prostatectomy. METHODS The biochemical test results and clinical data of 193 patients who underwent radical prostatectomy for the first time from January 2015 to December 2020 were retrospectively collected. The PNI value of peripheral blood within 1 wk before surgery was calculated, and during the follow-up period, prostate-specific antigen >= 0.2 ng/mL was considered to have biological recurrence. The receiver operating characteristic (ROC) curve was used to calculate the optimal critical value and area under the curve (AUC) of the patients. According to the critical value, the progression-free survival of the high PNI group and low PNI group was compared. The independent influencing factors of the patients' prognosis were obtained by the Cox proportional hazards regression model. RESULTS The non-biological recurrence rates at 1, 3, and 5 years were 92.02%, 84.05%, and 74.85%, respectively. The optimal critical value for PNI to predict biological recurrence was 46.23, and the AUC was 0.789 (95% confidence interval: 0.651-0.860; P < 0.001). The sensitivity and specificity were 82.93% and 62.30%, respectively. In accordance with the optimal critical value of the ROC curve (46.23), 193 patients were further divided into a high PNI group (PNI <= 46.23, n = 108) and low PNI group (PNI > 46.23, n = 85). The incidence of postoperative complications in the high PNI group was lower than that in the low PNI group (21.18% vs 38.96%). Kaplan-Meier survival analysis showed that the overall survival rate at 5 years in the low PNI group was 87.96% (13/108), which was lower than that in the high PNI group (61.18%, 33/85; P < 0.05). Low PNI [hazard ratio (HR) = 1.74; P = 0.003] and positive incisal margin status (HR = 2.14; P = 0.001) were independent predictors of biological recurrence in patients with high/extremely high-risk PCa. CONCLUSION The PNI has predictive value for the prognosis of patients with high/extremely high-risk PCa, and is an independent prognostic factor. Patients with low PNI value have a shorter time of non-biological recurrence after prostatectomy. It is expected that the combined prediction of other clinicopathological data will further improve the accuracy and guide postoperative adjuvant therapy to improve the quality of prognosis.