Aims In intensified insulin therapy, the recent development of short-acting insulin analogues with a very rapid onset of action forces a new discussion in terms of the optimal injection-meal interval. This study evaluated prandial glycaemia in patients with Type 1 diabetes following the subcutaneous injection of soluble human insulin (HI) and the insulin analogue insulin aspart (IAsp) at different injection-meal intervals and investigated whether administration of IAsp after the meal might provide satisfactory metabolic control. Methods In a randomized, double-blind, double-dummy, four-period crossover study, 20 Type 1 diabetic patients were investigated. Prandial insulin was administered 15 min before the start of the meal (HI(-15min)), immediately before the meal (HI(0min); IAsp((0min))) and 15 min after the start of the meal (IAsp((+15min))). Results Plasma glucose excursions from baseline levels during the 4 h (PG(exc)) were highest with HI(0min) (17.9 mmol.l(-1).h; P < 0.05 vs. other treatments) and were not statistically different for HI(-15min), IAsp((0min)) and IAsp((15min)) (13.6, 11.9 and 14.2 mmol.l(-1).h, respectively). Maximum concentration of plasma glucose (PG(max)) was lowest with IAsp((0min)) (11.2 mmol/l; P < 0.05 vs. other treatments). PG(max) was comparable with HI(-15min), HI(0min) and IAsp((+15min)) (13.3, 14.1 and 13.2 mmol/l, respectively). Conclusions With regard to prandial glycaemia IAsp((+15min)) is as effective as HI(- 5min) and superior to HI(0min). Thus, post-prandial dosing of the insulin analogue IAsp offers an attractive and feasible therapeutic option for well-controlled patients with Type 1 diabetes mellitus.