Abstract STUDY QUESTION Do the Edinburgh Selection Criteria correctly identify female cancer patients under the age of 18 who are at risk of premature ovarian insufficiency (POI) as candidates for ovarian tissue cryopreservation (OTC)? SUMMARY ANSWER Patient assessment using these criteria accurately identifies those at risk of POI, who can be offered OTC and future transplantation as a means of fertility preservation. WHAT IS KNOWN ALREADY Treatment for childhood cancer can have adverse consequences on future fertility; at the time of diagnosis, fertility risk assessment should be undertaken in order to identify patients to whom fertility preservation should be offered. The Edinburgh selection criteria, based on planned cancer treatment and patient health status, are utilized to identify those at high risk and therefore eligible for OTC. However, this procedure is not without risk and there are few data on the efficacy of the procedure in prepubertal patients. As such, long-term follow-up of reproductive outcomes is necessary, to ensure that OTC is being offered appropriately. STUDY DESIGN, SIZE, DURATION Cohort study encompassing all females diagnosed with cancer under the age of 18 in South East Scotland, from 1 January 1996 to 30 April 2020. Patients were followed up for reproductive outcomes to assess for diagnosis of POI. PARTICIPANTS/MATERIALS, SETTING, METHODS A total of 638 eligible patients were identified; patients under the age of 12 or deceased before the age of 12 were excluded from the study, leaving a study population of 431 patients. Electronic records were reviewed for reproductive function, assessed by current menstruation, pregnancy (in the absence of POI diagnosis), reproductive hormone measurements, pubertal progression, or diagnosis of POI. Patients on hormonal contraception (other than for treatment of POI or panhypopituitarism with no history of gonadatoxic treatment) were excluded from analysis (n = 9). Analysis on remaining 422 patients was carried out using the Kaplan–Meier methods, with POI as the defined event, and Cox proportional hazards model. MAIN RESULTS AND THE ROLE OF CHANCE In the study population of 431 patients, median ages at diagnosis and analysis were 9.8 and 22.2 years, respectively. Reproductive outcomes were unavailable in 142 patients; the assumption was made that these patients did not have POI, but a subanalysis excluding these patients was also performed. Of the 422 patients aged >12 at analysis and not taking hormonal contraception, OTC was offered to 37 patients and successfully performed in 25 patients. Of the 37 patients offered OTC (one at time of relapse), nine (24.3%) developed POI. Of the 386 not offered OTC, 11 (2.9%) developed POI. The probability of developing POI was significantly higher in those offered OTC (hazard ratio [HR] 8.7 [95% CI 3.6–21]; P < 0.0001), even when those patients with unknown outcomes were excluded from the analysis (HR 8.1 [95% CI 3.4–20]; P < 0.001). All patients offered OTC who developed POI did so after treatment for primary disease; in those not offered OTC, five patients (45.5%) developed POI after treatment for disease relapse. LIMITATIONS, REASONS FOR CAUTION A significant number of patients had unknown reproductive outcomes; many of these patients were engaged in ongoing follow-up but did not have documented reproductive assessment. This may have introduced bias to the analysis and highlights the need for reproductive follow-up as part of routine cancer aftercare. In addition, the relatively young age of the patient population and short duration of follow-up in some cases demonstrates the need for ongoing follow-up of this cohort. WIDER IMPLICATIONS OF THE FINDINGS The prevalence of POI after childhood cancer is low, but the Edinburgh selection criteria remain a robust tool for selecting those at high risk at the time of diagnosis, to offer OTC appropriately. However, disease relapse necessitating more intensive treatments remains a challenge. This study additionally highlights the importance of routine assessment and documentation of reproductive status in haematology/oncology follow-up. STUDY FUNDING/COMPETING INTEREST(S) K.D. is supported by a CRUK grant (C157/A25193). This work was undertaken in part in the MRC Centre for Reproductive Health, (supported by MRC grant MR/N022556/1). R.A.A. has received consulting fees from Ferring and Roche Diagnostics; payment from Merck and IBSA for educational events; and laboratory materials from Roche Diagnostics. The other authors have no competing interests to declare. TRIAL REGISTRATION NUMBER N/A.