Aim: Nonfasting plasma glucose is claimed to be a better marker of diabetic control than fasting plasma glucose in type 2 diabetes. In this study we compared the efficacy and safety profile of two different intensive insulin treatment strategies in patients with uncontrolled type 2 diabetes despite using a twice-daily insulin regimen. Methods: We studied 60 insulin-treated patients who had uncontrolled type 2 diabetes. The study was a 6-month, open- label, randomised, parallel clinical trial conducted in two diabetes centres. The main end-points for analysis were weekly self-monitored blood glucose readings, HbA(1c) levels, total daily insulin dose, weight gain and the number of hypoglycaernic episodes. Results: The breakfast 2-h, lunch 2-h and dinner 2-h postprandial glucose values and pre-dinner glucose values were significantly lower in the Lispro group than the regular insulin group. The HbA(1c), value at the end of the study was significantly lower in the Lispro group (7.3 +/- 0.7%) compared with the regular insulin group (7.7 +/- 0.7%; P < 0.05). Mean insulin doses were similar in the treatment groups initially and at the end. There was a statistically significant increase in insulin dose in both groups from baseline to the end of the study (P < 0.05). Overall hypoglycaernia rates were comparably low and similar in both groups during the study. Conclusions: We have shown that mealtime insulin Lispro plus additional lunch and bedtime NPH insulin is superior to premeal regular insulin plus bedtime NPH insulin for overall glyceamic regulation with similar weight gain and comparable rates of hypoglycaernia. (C) 2003 Elsevier Ireland Ltd. All rights reserved.