OBJECTIVE To assess the potential of insulin lispro to limit the frequency of severe hypoglycaemia without compromising glycaemic control in a cohort of patients with type 1 diabetes who are at a high risk of severe hypoglycemia. Research design and methods An open-label, randomised, 12-month comparative crossover study of insulin lispro and regular human insulin was performed in 33 patients with type 1 diabetes with impaired hypoglycaemia awareness. The efficacy of each treatment was evaluated by glycaemic control (HbA(1c)), eight-point home blood glucose profiles, and the frequency and severity of hypoglycaemic episodes and quality of life. RESULTS Eighteen (55%) patients experienced one or more episodes of severe hypoglycaemia in the 48 weeks of study. There was a trend to a lower incidence of severe hypoglycaemia during treatment with insulin lispro in comparison with regular human insulin (55 vs 84 episodes, p=0.087). This resulted principally from a 47% lower incidence of nocturnal severe hypoglycaemia with insulin lispro (25 vs 47 episodes, p=0.11). The lower frequency of severe hypoglycaemia associated with insulin lispro was not explained by differences in glycated haemoglobin between insulin treatments (HbA(1c) 9.1% insulin lispro vs 9.3% regular human insulin). CONCLUSIONS In individuals with type 1 diabetes, who have impaired awareness of hypoglycaemia, treatment with insulin lispro may be associated with a lower incidence of severe hypoglycaemia manifested predominantly through less frequent nocturnal episodes. Insulin lispro may have a beneficial role in the management of patients with diabetes at risk of severe hypoglycaemia, although a larger study is required to confirm these findings.