The accuracy of blood pressure (BP) measurements in obese hypertensives is still controversial. We investigated whether a large sized cuff could be appropriate for BP measurements in patients with different arm circumferences. Fifty hypertensive patients (23 males and 27 females, mean age 54 +/- 12 years, mean weight 75 +/- 12 kg) underwent BP measurements under standardised conditions using both a standard cuff (23 x 12 cm) and a large cuff (34 x 15 cm). Ambulatory blood pressure monitoring (ABPM) was subsequently performed in each subject using a large cuff. The average of two measurements performed in the clinic was compared with the mean daytime measurements. Thirty-one subjects had an arm circumference of < 30 cm (group 1) and weighted 68 +/- 10 kg, while 19 (group 2) had a larger arm circumference (> or = 30 cm) and were overweight (86 +/- 8 kg). BP in group 1 was 149 +/- 23/91 +/- 12 mmHg with the standard cuff and 143 +/- 24/88 +/- 13 mmHg with the large cuff (P < 0.003). In group 2, the respective readings were 153 +/- 22/100 +/- 16 mmHg (standard cuff) and 144 +/- 19/93 +/- 16 mmHg (large cuff; P < 0.0004). These findings suggest that in overweight hypertensives, the use of the standard cuff results in overestimation of BP and that large cuffs should be used exclusively in this population. The use of a large cuff in lean patients has yet to be clarified.