AIMS: To compare insulin and glucose profiles during basal continuous subcutaneous infusion of a rapid-acting insulin analogue and once daily subcutaneous injection of a long-acting insulin analogue in Type 2 diabetes. METHODS: Twenty-one patients with Type 2 diabetes treated with oral glucose-lowering agents were randomized in this two-period crossover study to an equivalent 24-h dose of continuous subcutaneous infusion of insulin aspart and subsequently once-daily bedtime subcutaneous injection of insulin glargine, or vice versa, for eight consecutive days. Plasma profiles of insulin and glucose were recorded. RESULTS: On the last day of each treatment period, the area under the curve (AUC) for glucose was 10% lower on the continuous subcutaneous infusion regimen compared with the insulin injection regimen (P = 0.002). This was accomplished by a flat exogenous insulin infusion profile compared with a peaking profile with injected insulin (AUC was 74% higher after injection compared with pre-injection levels (P = 0.001)). During the last 6 days in each treatment period, the intra-subject variability of exogenous fasting insulin levels in the mornings was 41% lower during insulin infusion compared with insulin injection (P = 0.012). The corresponding intra-subject variability for fasting glucose only showed a tendency to be lower during infusion as compared to the injection regimen (28%; P = 0.104). Thirteen symptomatic-only or minor hypoglycaemic episodes were recorded during the entire infusion period compared with three episodes during the injection period. CONCLUSIONS: Basal continuous subcutaneous infusion of a rapid-acting insulin analogue improved plasma insulin (more flat insulin profile with a lower variability) and glucose (lower AUC) profiles compared with once-daily subcutaneous injection of a long-acting insulin analogue in Type 2 diabetes.