Aims/hypothesis: Insulin resistance is associated with abnormal microvascular function. Treatment with insulin sensitisers may provoke oedema, suggesting microvascular effects. The mechanisms underlying the peripheral oedema observed during glucose-lowering treatment with thiazolidinediones are unclear. Therefore we examined the effect of pioglitazone on microvascular variables involved in oedema formation. Methods:: Subjects (40-80 years) with type 2 diabetes and on insulin were randomised to 9 weeks of pioglitazone therapy (30 mg/day; n=14) or placebo (n=15). The following assessments were performed at baseline and 9 weeks: microvascular filtration capacity; isovolumetric venous pressure; capillary pressure; capillary recruitment following venous or arterial occlusion; postural vasoconstriction; and maximum blood flow. A number of haematological variables were also measured including vascular endothelium growth factor (VEGF), IL-6 and C-reactive protein (CRP). Results: Pioglitazone did not significantly influence any microcirculatory variable as compared with placebo (analysis of covariance [ANCOVA] for microvascular filtration capacity for the two groups, p=0.26). Mean VEGF increased with pioglitazone (61.1 pg/ml), but not significantly more than placebo (9.76 pg/ml, p=0.94).HbA 1c levels and the inflammatory markers IL-6 and CRP decreased with pioglitazone compared with placebo (ANCOVA: p=0.009, p=0.001 and p=0.004, respectively). Conclusions/interpretation: Pioglitazone improved glycaemic control and inflammatory markers over 9 weeks but had no effect on microcirculatory variables associated with oedema or insulin resistance in type 2 diabetic patients treated with insulin.