PREDICTIVE (Predictable Results and Experience in Diabetes through Intensification and Control to Target: An International Variability Evaluation) is a large, multi-national, open-label, prospective, observational study assessing the safety and efficacy of insulin detemir in clinical practice. A total of 20,531 patients with type 1 or 2 diabetes from 11 countries were prescribed insulin detemir and followed up after a mean of 14.4 weeks. The primary endpoint was incidence of serious adverse drug reactions (SADRs), including major hypoglycaemia. Secondary endpoints were: haemoglobin A(1c) (HbA(1c)), mean self-monitored fasting glucose, within-patient fasting glucose variability and body weight change. Two hundred and fourteen patients (1%) reported SADRs, including major hypoglycaemia. The incidence of major hypoglycaemic episodes was reduced from 3.0/patient-year at baseline to 0.7/patient-year at follow-up in type 1 patients (p < 0.0001), and from 0.8 to 0.1/patient-year in type 2 patients (p < 0.0001). Insulin detemir improved glycaemic control in type 1 and type 2 patients, with reductions in mean HbA(1c) (0.5% and 0.9%, respectively, p < 0.0001 for both), fasting glucose (1.7 and 2.6 mmol/l, p < 0.0001 for both) and within-patient fasting glucose variability (0.7 and 0.5 mmol/l, p < 0.0001 for both). There was a small decrease in mean body weight in both type 1 and 2 patients (-0.1 kg, p < 0.01 and -0.4 kg, p < 0.0001 respectively). Insulin detemir was used once- or twice-daily in 49% and 50% of type 1 patients, and 77% and 23% of type 2 diabetes patients, respectively. The 14-week observations from PREDICTIVE support clinical trial data showing that insulin detemir improves glycaemic control, with a lowered risk of hypoglycaemia and no weight gain.