We studied 20 patients with moderate to severe obesity (body mass index 37 +/- 8 kg/m2) with obese arms (arm circumference 37 +/- 4 cm). Their blood pressure was measured directly in the brachial artery in one arm and simultaneously indirectly in the other arm using either a large standard cuff (rubber bag 12 x 35 cm) or a new cuff (the Tricuff, Pressure Group AB, Stockholm) containing three rubber bags of different sizes, which automatically selected the appropriately sized bag in relation to arm circumference. Both cuffs showed a significant overestimation of the diastolic blood pressure (standard cuff 13 mm Hg, P < .001, Tricuff 6 mm Hg, P < .01). The error of the standard cuff was significantly greater than that of the Tricuff (P < .001). The differences in systolic blood pressure between the intraarterially and the indirectly measured results were small (0.2 to 3 mm Hg) and not statistically significant with either cuff. The new Tricuff offered an advantage as compared to a large standard cuff in terms of a lesser overestimation of the diastolic blood pressure. In practical terms, this should lead to a reduction in the overestimation of diastolic hypertension in obese individuals.