The greatest benefit from novel glycaemic interventions in diabetes may be observed in people prone to hypoglycaemia. Phase 3 trials, however, usually specifically exclude people with recurrent severe hypoglycaemia or impaired hypoglycaemia awareness from participation. However, in clinical practice new glycaemic interventions e.g. insulins are often implemented first in those with problematic hypoglycaemia. Therefore, trials dedicated to study these people are needed to provide the evidence. The design should allow for comparison of rates of hypoglycaemia at comparable glycaemic control and should seek to reduce the risk of major random effects from individuals with very high rates of hypoglycaemia. Rates of hypoglycaemia are usually subject to a large and unpredictable general study effect and formal power calculations are difficult to perform. Since many hypoglycaemia-prone people have impaired hypoglycaemia awareness the major burden of hypoglycaemia may be asymptomatic. These events should be captured by continuous glucose monitoring (CGM).