Background: Hypoglycaemia, especially nocturnal, remains a main limiting factor of achieving good glycemic control in type 1 diabetes. The long-acting insulin analog degludec reduces the risk of nocturnal hypoglycemia in patients with type 1 diabetes but is not studied in patients specifically prone to severe nocturnal hypoglycemia. The study aims to investigate whether insulin degludec in comparison with insulin glargine U100 reduces the risk of nocturnal hypoglycaemia in patients with previous severe nocturnal hypoglycaemia. Methods: Investigator-initiated, prospective, randomized, open, blinded endpoint (PROBE), multicenter, cross-over study. A one-year plus one-year treatment period was specified. Each one-year treatment period consisted of a three-month run-in/cross-over period followed by a nine-month maintenance period. A total of 149 adult patients with type 1 diabetes and at least one episode of nocturnal severe hypoglycemia during the preceding two years were included. Participants were randomized 1:1 to basal- bolus therapy with insulin degludec and insulin aspart or insulin glargine U100 and insulin aspart. The primary endpoint was number of nocturnal symptomatic hypoglycemic episodes during the maintenance period, analysed by intention-to-treat. Findings: Treatment with insulin degludec resulted in a 28% (95%CI: 9-43; p=0·02) and 37% (95%CI: 16-53; p=0·002) relative rate reduction (RRR) of level 1 (≤3·9 mmol/L) and level 2 (≤3·0 mmol/L) nocturnal (00:00 to 05:59) symptomatic hypoglycaemia, respectively, compared to insulin glargine U100. Similar results were demonstrated when defining night-time from 23:00 to 06:59 with a 28% (95%CI: 4-45; p=0·01) and 34% (95%CI:17-48; p<0·001) RRR for level 1 and 2 hypoglycaemia, respectively. Interpretation: Patients with type 1 diabetes prone to nocturnal severe hypoglycaemia have lower rates of nocturnal symptomatic hypoglycaemia with insulin degludec as compared with insulin glargine U100.