Insulin aspart (IAsp), is a rapid-acting analogue of human insulin (HI), for use in the meal related treatment of diabetes mellitus. The degree of glycaemic control achieved by IAsp in comparison with HI after algorithm-driven dose optimisation was tested over 3 months. The prospective, multicentre, randomised, open-label study with parallel groups was performed in 48 centres in 11 countries and included 423 basal-bolus treated patients with Type 1 diabetes. Main outcome measures were blood glucose control assessed by HbA1c, nine-point self-monitored blood glucose profiles, insulin dose, quality of life, hypoglycaemia and adverse events. An algorithm-driven increase occurred in the dose and number of daily injections of basal insulin, particularly in the IAsp group. After 12 weeks of treatment, HbA1c was significantly lower in IAsp compared to HI treated subjects by 0.17 (95% CI 0.30-0.04) (P<0.05). Comparison of the blood glucose profiles showed lower blood glucose levels with IAsp after breakfast (mean 8.4 vs 10.1 mmol/l; P<0.0001) and dinner (8.2 vs 9.3 mmol/l; P<0.01). There were no differences between treatments in the incidence of hypoglycaemic episodes or in the adverse event profiles. The WHO Diabetes Treatment Satisfaction Questionnaire score for perceived hyperglycaemia was lower with Iasp (P=0.005), and patients found the insulin aspart treatment more flexible (P=0.022). The current study underlines the need for optimising the basal insulin regimen in order to take full advantage of the pharmacodynamics of IAsp.