Abstract Study question What is the prevalence, and which factors are associated with hypogonadism in infertile men with non-obstructive azoospermia (NOA)? Summary answer The prevalence of biochemical hypogonadism among NOA males is significant. The condition is associated with testicular volume, estradiol levels, paternal age, and testicular histopathology results. What is known already NOA is a severe and irreversible condition that accounts for approximately 60% of azoospermia cases, primarily associated with intrinsic testicular deficiencies of various causes. NOA men can present signs of hypogonadism, associated or not with elevated circulating FSH levels; however, the exact prevalence is ill-reported. There is also a lack of data on the clinical factors associated with hypogonadism. We, therefore, estimated the prevalence of biochemical hypogonadism using real-world data and investigated its relationship with clinical factors. Study design, size, duration An observational cohort study was conducted, including 767 consecutive NOA patients with primary spermatogenic failure seeking fertility treatment at a University-affiliated tertiary center for male reproductive health between January 2014 and September 2021. Biochemical hypogonadism was defined as total testosterone (T) levels <350 ng/dL, measured by chemiluminescence immunoassay from a venous sample, collected in the morning, from 8:00 to 10:00 am, and confirmed on a second analysis at least one week apart. Participants/materials, setting, methods All included patients (age range: 23-55 years) had complete clinical, hormonal, genetic, seminal, and histopathology data and were naïve concerning previous sperm retrieval attempts. Patients with NOA due to hypogonadotropic hypogonadism and those using hormonal therapy were excluded. Prevalence rates were computed by Goodman’s method. The relationship between clinical factors and hypogonadism was assessed by logistic regression analysis. We also assessed the frequency of patients with high and within-range FSH levels in the cohort. Main results and the role of chance NOA patients with biochemical hypogonadism represented 80.3% (95% confidence interval [CI] 77.3–43.7) of the studied population. The median (interquartile range [IQR]) of T and FSH levels in hypogonadal men were 276 ng/dL (233.0-303 ng/dL) and 12.3 IU/L (8.9-18.2). Among them, hypergonadotropic (FSH>12 IU/L) and normogonadotropic (within-range FSH levels 1.5-12.0 IU/L) patients were 52.1% (95% CI 48.1-56.0) and 47.8% (43.9-51.8) of studied individuals, respectively. Logistic regression analyses revealed that testicular volume (p < 0.0001) and estradiol levels (p = 0.0005) had a significant inverse relationship with hypogonadism, whereas paternal age (p = 0.02) showed a positive relationship. Moreover, testicular histopathology was a relevant predictor (p = 0.003), with an inverse relationship between hypogonadism and germ cell maturation arrest and a positive relationship between hypogonadism and hypospermatogenesis or Sertoli cell-only. Using these variables, a model constructed to predict hypogonadism had an area under the ROC curve of 0.72. Limitations, reasons for caution Limitations include possible unmeasured confounding factors and the intra- and inter-assay electrochemiluminescence method variability. Another limitation is that the study center is a referral facility for azoospermic patients; thus, hypogonadism prevalence rates reported herein might not represent that of the general population of NOA males attending general fertility centers. Wider implications of the findings Efforts are needed to provide adequate care for this population due to its association with poor quality of life. Moreover, as recent evidence has suggested that hormonal therapy to boost T production might increase sperm retrieval success rates, studies investigating hormonal therapy in such patients may be warranted. Trial registration number Not applicable