Abstract Study question Is there an association between pre-operative FSH levels and successful outcomes following Surgical Sperm Retrieval (SSR), and subsequent fertility treatment? Summary answer Higher levels of pre-operative FSH are associated with unsuccessful SSR, and, in the cases of successful retrieval, sperm less likely to produce healthy live births. What is known already Surgical Sperm Retrieval (SSR) techniques combined with Intracytoplasmic Sperm Injection (ICSI) offer men with azoospermia a chance to father biological children. However, the variable success rates of these physically, psychologically, and financially costly procedures highlights the potential value of pre-operative predictors of success. A literature review indicated serum FSH a promising pre-operative prognostic factor. However, only very few papers investigated pregnancy outcomes as an endpoint. Study design, size, duration An extensive literature review was undertaken examining literature between 2015-2020. This led to determination of the serum cut-off level to apply to the different groups’ outcomes. The study was a retrospective clinical analysis of SSR offered by a university Hospital between 2014-2021. Sample size was 118 (mainly Testicular Sperm Aspiration (TESA) and Percutaneous Epididymal Sperm Aspiration (PESA) patients. Patients with Klinefelter’s (KF) (N = 3) were excluded from analysis due to the genetic nature of their azoospermia. Participants/materials, setting, methods Databases from the University Hospital's Theatres and Andrology Department provided data on patients who underwent SSR 2014-present. A retrospective evaluation of physical and electronic notes identified preoperative hormone profiles and other potential influencers. Pregnancy outcomes were ascertained from the ICSI and Frozen Embryo transfer (FET) databases. Using a serum FSH cut-off of 15IU/mL (chosen based on previous literature review) data was analysed for differences between the groups’ outcomes. ROC analysis generated potential alternate cut-offs. Main results and the role of chance Using the initial 15IU/mL cut-off analysis demonstrated significantly more patients had unsuccessful SSR and sperm that did not produce healthy live birth where FSH was ≥15IU/mL at a 95% confidence level. The ROC curves for both SSR and pregnancy outcomes also proved significant – suggesting various potential cut-offs from our data. One potentially valuable cut-off for identifying men unlikely to produce sperm capable of producing a healthy live birth following all treatment is 14.0IU/mL (FSH above this predicting failure in pregnancy outcomes) (sensitivity 40.8%, specificity 90.5%), i.e., above this level the model predicts 90.5% of men will not have success. The lower bound 95% CI for this curve was above 0.5, indicating the curve to be statistically significant at a 95% confidence level. Limitations, reasons for caution SSR is rare and a lack of documentation available for some patients resulted in a relatively small sample size. Our data comprises primarily TESE and PESA patients. Different surgeon technique or service standards may influence outcomes. Our findings did not consider KF patients, who may have favourable SSR outcomes. Wider implications of the findings SSR is a costly procedure financially and physically with variable success rates. Hopefully these findings can help counsel men as to their likelihood of success in having a healthy live birth following SSR and subsequent treatment, helping to reduce both a financial burden on services and emotional burden on patients. Trial registration number Not applicable