BACKGROUND: Non-obstructive azoospermia (NOA) is a prevalent cause of male infertility, characterized by the lack of sperm in the ejaculate due to impaired spermatogenesis. Accurate prediction of testicular sperm extraction (TESE) outcomes is pivotal for counseling and managing patients, yet remains challenging due to variability in clinical presentations. This study aimed to establish a predictive model for TESE outcomes in patients with NOA. METHODS: We retrospectively analyzed 425 patients who visited the Andrology Outpatient Clinic of the First Affiliated Hospital of Soochow University between January 2010 and January 2024. Of these, 216 had positive sperm retrieval, and 209 had negative outcomes. We compared testicular volume, reproductive hormone levels, and other clinical parameters between two groups. Multivariate logistic regression was used to identify independent risk factors that were used to establish a predictive nomogram. A calibration curve was used to evaluate the model's fit, whereas receiver operating characteristic (ROC) curves and decision curve analysis (DCA) were used to assess diagnostic effectiveness and net benefit. RESULTS: The differences were significant in serum follicle-stimulating hormone (FSH) (P<
0.001), body mass index (P=0.04), elastase levels (P=0.005), and testicular microlithiasis levels (P=0.005) between the groups. Multivariate regression identified FSH (P<
0.001), testicular volume (P<
0.001), and testosterone levels (P=0.003) as independent risk factors for TESE outcomes. FSH levels were negatively correlated [odds ratio (OR) =0.905, 95% confidence interval (CI): 0.876-0.935, P<
0.001], while testicular volume (OR =1.453, 95% CI: 1.328-1.591, P<
0.001) and testosterone (OR =1.326, 95% CI: 1.098-1.601, P=0.003) were positively correlated. Nomogram based on these factors showed a good fit with an area under the ROC curve of 0.879. The DCA plot demonstrated substantial clinical benefits. CONCLUSIONS: In patients with NOA, low testicular volume, low testosterone levels, and high FSH levels were independent risk factors for unsuccessful TESE. The predictive nomogram provided excellent predictive power for positive TESE outcomes.