This is the final version of the article. Available from Springer Verlag via the DOI in this record. ; AIMS/HYPOTHESIS: The finding that patients with diabetes due to potassium channel mutations can transfer from insulin to sulfonylureas has revolutionised the management of patients with permanent neonatal diabetes. The extent to which the in vitro characteristics of the mutation can predict a successful transfer is not known. Our aim was to identify factors associated with successful transfer from insulin to sulfonylureas in patients with permanent neonatal diabetes due to mutations in KCNJ11 (which encodes the inwardly rectifying potassium channel Kir6.2). METHODS: We retrospectively analysed clinical data on 127 patients with neonatal diabetes due to KCNJ11 mutations who attempted to transfer to sulfonylureas. We considered transfer successful when patients completely discontinued insulin whilst on sulfonylureas. All unsuccessful transfers received ≥0.8 mg kg(-1) day(-1) glibenclamide (or the equivalent) for >4 weeks. The in vitro response of mutant Kir6.2/SUR1 channels to tolbutamide was assessed in Xenopus oocytes. For some specific mutations, not all individuals carrying the mutation were able to transfer successfully; we therefore investigated which clinical features could predict a successful transfer. RESULTS: In all, 112 out of 127 (88%) patients successfully transferred to sulfonylureas from insulin with an improvement in HbA1c from 8.2% (66 mmol/mol) on insulin, to 5.9% (41 mmol/mol) on sulphonylureas (p = 0.001). The in vitro response of the mutation to tolbutamide determined the likelihood of transfer: the extent of tolbutamide block was 73% did transfer successfully. The few patients with these mutations who could not transfer had a longer duration of diabetes than those who transferred successfully (18.2 vs 3.4 years, p = 0.032). There was no difference in pre-transfer HbA1c (p = 0.87), weight-for-age z scores (SD score; p = 0.12) or sex (p = 0.17). CONCLUSIONS/INTERPRETATION: Transfer from insulin is successful for most KCNJ11 patients and is best predicted by the in vitro response of the specific mutation and the duration of diabetes. Knowledge of the specific mutation and of diabetes duration can help predict whether successful transfer to sulfonylureas is likely. This result supports the early genetic testing and early treatment of patients with neonatal diabetes aged under 6 months. ; This research was supported by the National Institute for Health Research (NIHR) Exeter Clinical Research Facility, the Wellcome Trust and the European Research Council. FMA is a European Research Council Advanced Investigator and holds a Royal Society/Wolfson Merit Award. SE is a Wellcome Trust Senior Investigator. ATH is a Wellcome Trust Senior Investigator and a NIHR Senior Investigator.