Insulin resistance is associated with endothelial dysfunction in type 2 diabetes mellitus, which can lead to impaired vascular reactivities of both systemic and cerebral circulations. Appropriate 'correction' of vascular reactivity results for non-endothelium-dependent systemic effects avoids misinterpretation of endothelial function. Therefore, we 'corrected' vascular reactivity results and explored the potential correlations between systemic vascular reactivity, cerebrovascular reactivity and insulin resistance. In 34 patients, 'systemic vascular reactivity' was assessed by quantifying reactive hyperaemia. Cerebrovascular reactivity was assessed by quantifying changes in cerebral blood flow velocity during hypercapnia. To minimize the influence of non-endothelium-dependent systemic effects on vascular reactivity results, 'corrected systemic vascular reactivity' was calculated by normalizing systemic vascular reactivity using the measurements from the contralateral side; and cerebrovascular reactivity results were corrected by calculating percentage and absolute changes in cerebrovascular conductance index ('percent cerebrovascular conductance index' and 'delta cerebrovascular conductance index', respectively). Insulin resistance was estimated by homeostatic model assessment. Correlation between conventional cerebrovascular reactivity and systemic vascular reactivity was not significant. But correlations between 'corrected systemic vascular reactivity' and 'percent cerebrovascular conductance index' (r = 0.51; p = 0.002) and 'corrected systemic vascular reactivity' and 'delta cerebrovascular conductance index' (r = 0.50; p = 0.003) were significant. Among all vascular reactivity parameters, only 'delta cerebrovascular conductance index' was significantly correlated with homeostatic model assessment of insulin resistance (r = −0.38; p = 0.029). In conclusion, endothelial function in the systemic and cerebral circulations is moderately correlated, provided that vascular reactivity estimates are corrected for non-endothelium-dependent influences.