Objective: To evaluate the effect of rosiglitazone, an insulin sensitizer, on glycaemic control and insulin resistance in adolescents with type 1 diabetes mellitus (T1DM). Research design and methods: Randomized, double-blind, placebocontrolled crossover trial of rosiglitazone (4 mg twice daily) vs. placebo (24 wk each, with a 4 wk washout period). Entry criteria were diabetes duration >1 yr, age 10–18 yr, puberty (≥Tanner breast stage 2 or testicular volume >4 mL), insulin dose ≥1.1 units/kg/day, and haemoglobin A1c (HbA1c) >8%. Responses to rosiglitazone were compared with placebo using paired t-tests. Results: Of 36 adolescents recruited (17 males), 28 completed the trial. At baseline, age was 13.6 ± 1.8 yr, HbA1c 8.9 ± 0.96%, body mass index standard deviation scores (BMI-SDS) 0.94 ± 0.74 and insulin dose 1.5 ± 0.3 units/kg/day. Compared with placebo, rosiglitazone resulted in decreased insulin dose (5.8% decrease vs. 9.4% increase, p = 0.02), increased serum adiponectin (84.8% increase vs. 26.0% decrease, p < 0.01), increased cholesterol (+0.5 mmol/L vs. no change, p = 0.02), but no significant change in HbA1c (-0.3 vs. -0.1, p = 0.57) or BMISDS (0.08 vs. 0.04, p = 0.31). Insulin sensitivity was highly variable in the seven subjects who consented to euglycaemic hyperinsulinaemic clamps. There were no major adverse effects attributable to rosiglitazone. Conclusion: The addition of rosiglitazone to insulin did not improve HbA1c in this group of normal weight adolescents with T1DM.