BACKGROUND: Hypoglycaemia, a common complication of insulin-treated diabetes, especially type 1 diabetes (T1D), is caused by impaired counterregulatory increases in plasma glucagon, leading to decreased hepatic glucose production. Alpha-cells in the pancreatic islets, which have sulfonylurea-sensitive ATP-regulated potassium (KATP) channels, secrete glucagon at low plasma glucose levels in response to increased electrical activity. Activation of KATP channels inhibits alpha-cell electrical activity and consequently reduces glucagon secretion. Excessive secretion of somatostatin from neighbouring delta-cells (which are sensitive to the SGLT2 inhibitor dapagliflozin) also causes inappropriate paracrine inhibition of glucagon secretion at low glucose levels. We hypothesise that increased KATP channel activity and/or intra-islet somatostatin may contribute to the glucagon secretion defects seen in T1D and that these issues can be addressed with sulfonylureas or SGLT2 inhibitors. METHODS: We first tested this hypothesis in vitro by comparing glucagon secretion in response to increasing concentrations of the sulfonylurea tolbutamide-used to titrate KATP channel activity-in isolated human islets from donors with T1D and healthy donors without a history of diabetes. Next, we conducted a single-centre, randomised, triple crossover, phase 2a superiority trial in people with T1D to evaluate the effectiveness of oral glibenclamide (0.3, 0.6, and 3 mg/day, over 14-18 days per dose) or an acute single dose of dapagliflozin (10 mg) in restoring glucagon counter-regulation. This was assessed using a hyperinsulinaemic-hypoglycaemic clamp in 12 male participants with T1D, with responses from 6 healthy controls (4 males and 2 females) serving as a reference. The primary endpoint was the change in plasma glucagon levels during hypoglycaemia from baseline (no medication) to post-treatment (glibenclamide or dapagliflozin). All participants were included in the analyses. The study is registered with ISRCTN (58098350). FINDINGS: In islets from 9 healthy male donors, lowering glucose from 6 to 1 mM (to emulate hypoglycaemia) stimulated glucagon secretion by 120%. This effect was not observed in islets from 7 donors with T1D (4 males and 3 females). In these islets, a low concentration (10 μM) but not a high (100 μM) concentration of the sulfonylurea tolbutamide stimulated glucagon release by 70%. During the trial, insulin-induced hypoglycaemia (plasma glucose: <3 mM), in the absence of glibenclamide or dapagliflozin, produced a >700% increase in circulating glucagon in the control group, but only a 54% increase in participants with T1D. The latter value fell to <10% in the 8 participants with low circulating C-peptide levels (≤6 pM). In these participants, glibenclamide (0.3-0.6 mg/day for 14-18 days) raised circulating glucagon during hypoglycaemia by 125-150%; this correlated with reduced glucose infusion requirements, reflecting stimulation of endogenous glucose production. No effect on either parameter was seen with 3 mg/day glibenclamide. These increases in circulating glucagon were associated with better glucose control, as measured by continuous glucose monitoring during the last 7 days of each treatment period. Similar effects on plasma glucagon secretion and glucose infusion rates were observed with the SGLT2 inhibitor dapagliflozin. INTERPRETATION: Enhanced KATP channel activity in alpha-cells and/or increased paracrine inhibition by somatostatin impairs the counterregulatory glucagon response in T1D. These issues can be partially corrected by glibenclamide or dapagliflozin. We suggest that these agents be further tested as adjuvants to insulin therapy in T1D to help reduce the occurrence of iatrogenic hypoglycaemia and the risks it poses. FUNDING: Supported by The Leona M. and Harry B. Helmsley Charitable Trust. The funding body did not influence the study's design, data analysis, or interpretation.